TWI687686B - Method, apparatus and system for examining biological epidermis - Google Patents

Abstract

The present application discloses an examining biological epidermis method for providing quantized parameters with regard to an examining area. The method comprises: taking a photo showing one or more color samples and the examining area by a camera module; searching and locating the color samples on the photo according to multiple predetermined characteristic values of color samples; correcting the photo according to differences between the color samples and the characteristic values; searching and locating the examining area on the corrected photo according to predetermined examining conditions; and calculating the quantized parameters.

Description

Method, device and system for detecting biological epidermis

The present application relates to a method for detecting biological epidermis, in particular to a method for photographic detection of biological epidermis using color block samples.

Traditional biological epidermal layer testing is usually used to detect wounds, scars, pigmentation and other defects, after the use of drugs or maintenance products. Due to the long hours of effectiveness of drugs or maintenance products, it usually takes days or even weeks to achieve their effects. Therefore, traditionally, photographs are used to preserve the condition of the epidermis before medication, and then the condition of the epidermis between medication and after medication is compared.

Please refer to FIG. 1, which is a photo 100 of the epidermis of the prior art. There is a flaw 110 in the photo 100. However, looking at the photo 100, the color and size of the defect 110 cannot be seen. What's more, because the light source may be different when taking pictures, the camera shooting parameters may be different, and the shooting angle may also be different. This makes it difficult to interpret the photos. Different interpreters may come to different interpretations. Moreover, the interpretation results cannot be quantified, and the effects of drugs or maintenance products cannot be accurately measured. Therefore, there is an urgent need for a method that can quantitatively present the results of epidermal layer detection in order to correctly evaluate the effects of drugs or maintenance products.

According to an embodiment of the present application, a biological epidermal layer detection method is provided for Provides a quantization parameter of the detection area, including: using a camera module to take a photo, wherein the photo contains one or more color patch samples and the detection area on the biological epidermis; according to the characteristics of the plurality of color patch samples predicted in advance Value, search and locate one or more of the color patch samples in the photo; correct the photo based on the difference between the feature values of the one or more color patch samples and the multiple color patch samples; according to the preset Detection conditions, search and locate the detection area in the corrected photos; and calculate the quantization parameters of the detection area.

According to an embodiment of the present application, a biological epidermal layer detection device is provided for providing quantization parameters of a detection area, including: a camera module for taking a photo, wherein the photo includes one or more color patch samples And the detection area on the biological epidermis; a memory module for storing a detection application, the photo and the characteristic values of a plurality of color patch samples; and a central processor connected to the camera module and the memory The body module is used to execute the following steps of the detection application: search and locate one or more of the color patch samples in the photo according to the previously predicted feature values of the plurality of color patch samples; according to one or more of the The difference between the characteristic values of the color patch samples and the plurality of color patch samples is used to correct the photo; the detection area is searched and located in the corrected photo according to the detection conditions preset in advance; and the quantization parameter of the detection area is calculated.

According to an embodiment of the present application, a biological epidermal layer detection system is provided for providing a quantitative parameter of a detection area, including: at least one color patch sample; and a biological epidermal layer detection device. The embodiment of the biological epidermal layer detection device is as described above.

According to the above embodiments, the present application provides a method, device, and system that can quantitatively present the results of epidermal layer detection, and can accurately evaluate the effects of drugs or maintenance products.

100‧‧‧Photo

110‧‧‧Defective or detection area

200‧‧‧Photo

210‧‧‧color sample

212‧‧‧Blue color block

214‧‧‧Red color block

216‧‧‧ green color block

300‧‧‧Epidermal layer detection system

310‧‧‧Color sample

320‧‧‧Camera module

330‧‧‧ I/O device bridge

340‧‧‧ screen module

350‧‧‧Memory module

360‧‧‧ CPU

400‧‧‧Epidermal layer detection method

410~490‧‧‧ steps

500‧‧‧Color block

501‧‧‧ First pixel

502‧‧‧ second pixel

503‧‧‧ third pixel

504‧‧‧ fourth pixel

512‧‧‧ vector

513‧‧‧ Vector

520‧‧‧ slope line

522‧‧‧Average

524‧‧‧ Vector

526‧‧‧Average

534‧‧‧ vector

600‧‧‧Photo

610‧‧‧Connect

620‧‧‧Color block sample

630‧‧‧Yellow color block

700‧‧‧Photo

712‧‧‧ color block sample

714‧‧‧ color block sample

722‧‧‧ virtual color sample

724‧‧‧ virtual color sample

800‧‧‧Epidermal layer detection method

810~880‧‧‧Step

910‧‧‧Projection position

913‧‧‧ First length

915‧‧‧ slope line

920‧‧‧Projection position

925‧‧‧ slope line

930‧‧‧Projection position

932‧‧‧Second length

935‧‧‧ slope line

950‧‧‧ Connection of slope line

1010‧‧‧Projection position

1015‧‧‧ slope line

1020‧‧‧Projection position

1025‧‧‧ slope line

1050‧‧‧Conic

FIG. 1 is a schematic diagram of a photograph of a defect in the epidermis of the prior art.

FIG. 2 is a schematic diagram of a photograph according to an embodiment of the present application.

FIG. 3 is a block diagram of an epidermal layer detection system according to an embodiment of the present application.

FIG. 4 is a schematic flow chart of one method of epidermal layer detection according to an embodiment of the present application.

FIG. 5A is a schematic diagram of a color patch detection point according to an embodiment of the present application.

FIG. 5B is a schematic diagram of color block correction slope according to an embodiment of the present application.

FIG. 5C is a schematic diagram of a two-dimensional corrected slope of a color patch according to an embodiment of the application.

6 is a schematic diagram of a photograph according to an embodiment of the present application.

7 is a schematic diagram of a photo correction according to an embodiment of the present application.

FIG. 8 is a schematic flow chart of a method for detecting an epidermal layer according to an embodiment of the present application.

9 is a schematic diagram of color correction according to an embodiment of the present application.

FIG. 10 is a schematic diagram of color correction according to an embodiment of the present application.

This application will describe some embodiments in detail as follows. However, in addition to the disclosed implementation exceptions, the present application can also be widely applied in other embodiments. The scope of this application is not limited by these embodiments, but is subject to the scope of subsequent patent applications. In order to provide a clearer description and enable those familiar with the art to understand the content of this application, the parts in the illustration are not drawn according to their relative sizes, and the ratio of some sizes to other related scales will be highlighted and appear Exaggerated, and irrelevant details have not been completely drawn, in order to make the illustration concise. In addition, among the steps shown in the flowcharts of this application, other steps not related to this application may be inserted. Unless there is a cause-and-effect dependency relationship, this application does not limit the execution order of each step.

Please refer to FIG. 2, which is a schematic diagram of a photo 200 according to an embodiment of the present application. The photo 200 includes at least one surface layer defect 110 and at least one color block sample 210. The latter is placed or attached near the surface layer defect 110 for correcting the color reference value of the photo 200 and used as a scale. In other words, the color, shape and size contained in the color patch sample 210 are known in advance.

In an embodiment, when the photo 200 is a black-and-white photo, the color patch sample 210 may be a color patch composed of a plurality of gray levels. For example, in addition to the conventional visible band camera, the photo may be taken by an infrared camera, or may be taken by an ultraviolet camera. The grayscale depth can be 8 bits or deeper. In another embodiment, when the photo 200 is a color photo, the color patch sample 210 may be a color patch composed of a plurality of colors. In the embodiment shown in FIG. 2, the color block sample 210 includes a blue color block 212, a red color block 214 and a green color block 216, which are exactly three primary colors. However, the present application does not limit that the color block sample 210 contains only the three primary colors. For example, the color block sample 210 may include CMYK cyan, magenta, yellow, and black four-color swatches.

After the camera shoots, the photo 200 can be stored as pixels represented by various color modes, and represented by a two-dimensional array of a first axis and a second axis. About the commonly used color mode, it can be the common RGB color mode, CMYK color mode, CIE L*a*b* color mode, index mode, gray scale mode, etc. These color modes can be converted to each other.

Please refer to FIG. 3, which is a block diagram of an epidermal layer detection system 300 according to an embodiment of the present application. The detection system 300 includes at least one color block sample 310 and a detection device. The detection device further includes a camera module 320, an input/output device bridge 330, a screen module 340, a central processing unit 360, and a memory module 350. For example, the detection device may It can be a smartphone, laptop, tablet or other form of computer.

As mentioned above, the camera module 320 can be used for digital cameras that capture visible light, infrared, and/or ultraviolet light. It can output the taken photo 200 to the input/output device bridge 330 as an input. The interface between the two can be the industry standard USB, SATA, ATA, PCI-E, PCI, SCSI and other interfaces, or it can be a dedicated interface. This application does not limit the interface between the two. The I/O device bridge 330 can output the photo 200 to the screen module 340 for viewing by the user. In one embodiment, the detection device may provide a preview function, that is, before the photo is stored, the user is allowed to preview the photo to be taken on the screen module 340.

The I/O device bridge 330 is connected to a central processing unit 360 for executing the operating system and testing applications, and controlling the testing device. The central processing unit 360 is connected to the memory module 360 for reading and executing the operating system and testing application programs stored in the non-volatile storage medium contained therein. The photo 200 taken by the camera module 320 can also be stored in the memory module 350. The detection application is composed of a plurality of commands and data, including previously known color sample characteristics and historical records, which can be used to correct the photo 200 to perform a detection method of quantitatively presenting the results of epidermal layer detection in order to Correctly evaluate the effect of the drug or maintenance product on the defect 110. In one embodiment, the detection application can be downloaded to the memory module 350 via a network module not shown in the figure. For example, the detection application can be downloaded from Apple’s App Store and/or Google’s Google Play online store.

Please refer to FIG. 4, which is a schematic flowchart of a method 400 for epidermal layer detection according to an embodiment of the present application. The detection method can be applied to the detection application program executed by the central processor 360 of FIG. 3, or the detection application program can be used to execute the epidermal layer detection method 400 Part of the steps.

As mentioned above, among the steps shown in the flowcharts of this application, other steps not related to this application may be inserted. Unless there is a cause-and-effect dependency relationship, this application does not limit the execution order of each step. The detection method 400 includes the following steps:

Step 410: Paste a color block sample near the detection area of the biological epidermal layer to be detected. The detection area includes at least one defect area to be detected, such as the defect 110 shown in FIG. 2. The color patch sample may be the color patch sample 210 shown in FIG. 2. When the organism to be detected is a cooperative target, this step can be performed using an automated machine. For example, when you want to detect a human face, you can ask the subject to be close to a face frame, and the robot arm will paste the color block sample on the human face. The face frame and the robot arm may be part of the detection device of FIG. 3, connected to the I/O device bridge 330 and controlled by the detection application executed by the central processor 360.

Step 420: Use a camera module to take a photo, the photo contains the color block sample and the detection area. The photo may be the photo 200 of FIG. 2, and the camera module may be the camera module 320 of FIG. 3.

Step 430: Search and locate the color block sample in the photo according to the characteristics of the color block sample predicted in advance. The characteristics of the color patch samples referred to here may include, but are not limited to, the color, size, shape of the color patch samples and the positional relationship between them. Since there are already many open source softwares that provide the functions of face recognition or object recognition in photos, this application believes that ordinary people in the art can use existing technology to search and locate the color block samples in the photos, so they will not be repeated here. . Then, the process may enter optional step 440, or directly enter step 450. Of course, if no color block samples can be found, the process returns to step 420.

Optional step 440: determine whether the color patch sample is qualified? This application can be packaged There are more sub-steps to judge whether the color patch sample is qualified. For example, it can be judged whether the color patch sample meets a certain size, for example, its total area is greater than a few pixels, or the area of each color is greater than a few pixels. In another embodiment, it can be determined whether the difference between the maximum value and the minimum value of the color in each color block sample is less than a certain range. When the difference is too large, it may indicate that the color patch has partially faded, or is affected by the shadow and cannot be used together. In a further embodiment, it can be determined whether the maximum value of each color color block that is not the color in the color block sample is greater than a certain range. For example, when the pixels of the red color block have an excessively large blue or green value, it may indicate that the red color block has partially faded, or is affected by shadows and cannot be used together.

When the judgment result in step 440 is qualified, the flow proceeds to step 450. When the judgment result is unqualified, the flow returns to step 420. The above steps 420-440 can be performed in the aforementioned preview mode. When the user previews, the detection application can perform the calculation results of these steps on the current preview screen to determine whether the preview screen is qualified. When the preview screen is qualified, the user can save the photo. Otherwise, the user will not be allowed to save the photo and perform the following steps.

Step 450: Perform color correction on the photo according to the difference between the characteristics of the color patch sample and the color patch sample. The application may include more sub-steps to perform color correction. In an embodiment, each color value of the color patch sample may be averaged to obtain a difference value from the color value feature of the color patch sample. However, each pixel of the photo is compensated corresponding to the difference to obtain a corrected photo. In later sections of this application, color correction will be explained further.

Step 460: Search in the corrected photos according to the preset detection conditions And locate at least one of the detection areas. The detection conditions referred to herein may include, but are not limited to, the defects, or the color, size, and shape of the detection area and the positional relationship between them. Since there are already many open source softwares that provide functions for face recognition or object recognition in photos, this application believes that ordinary people in the art can use existing technology to search for and locate irregularly shaped defects or detection areas in the photos, so it is not Repeat them again. In one embodiment, the blemishes may include, but are not limited to, red and swollen wounds, deep black scars, or surface pigmented areas.

Optional step 470: Calculate the size and/or area of the detection area based on the size recorded in the characteristics of the color patch sample and the size of the corrected photo of the color patch sample. Since the color patch sample features include the size of each color patch, the size of the detection area can be estimated based on how many pixels the color patch occupies in the photo. The area of the detection area can also be estimated based on the number of pixels occupied by the detection area. The quantized parameters such as the size or area of the defect or detection area can be used in step 480 to compare with the historical record of the detection area to obtain a comparison result.

Optional step 480: Calculate the average color, color extremum and other parameters of the detection area. Since the parameters to be detected in each detection zone may be different, the quantized parameters referred to in this step include but are not limited to the average color, color extremum, size, area, etc. of the detection zone.

Optional step 490: Compare the parameter with the historical record of the detection area to obtain a comparison result. Because the aforementioned parameters have been quantified, it can be easily compared with historical records to obtain comparison results such as darker or lighter colors, larger or smaller sizes, area enlargement or reduction, and the proportion of improvement or deterioration over time. .

Please refer to FIG. 5A, which is a schematic diagram of a color patch detection point according to an embodiment of the present application. The color patch 500 may be a rectangle, including four corner pixels 501-504. The detection application can use the difference between these four corner pixels 501 and 504 to calculate the correction line. A person of ordinary skill in the art may understand that although FIG. 5A uses a rectangle as an example, the present application is not limited to a rectangle and its four corner points. Any point among hexagons, triangles, or other shapes can be used as an example.

In FIG. 5A, the first pixel 501 and the second pixel 502 can form a vector 512, and the first pixel 501 and the third pixel 503 can form another vector 513. The vector 512 and the vector 513 are perpendicular to each other, and after correction, they can be parallel to the straight axis and the horizontal axis of the photo 200, respectively. A person of ordinary skill in the art can understand the formation of the vectors 524 and 534, so the details are not repeated here. Each vector may have an error in the corresponding color. This error may come from the distance from the light source, or whether it fell into the shadow. In an ideal situation, the light source is located at infinity above the color patch sample. Then the pixel values of the four pixel points 501-504 should be the same or the error should be less than a predetermined range. However, when the light source is not in the ideal position, the pixel values of these four pixels 501-504 will change.

Please refer to FIG. 5B, which is a schematic diagram of a color block correction slope according to an embodiment of the present application. Taking vector 513 as an example, the color value corresponding to the first pixel point 501 on the right side is larger, and the color value corresponding to the third pixel point 503 on the left side is smaller, and both have a slope line 520. In step 450 of FIG. 4, the slope line 520 may be corrected or equalized to an average line 522 or another average line 526. The average line 522 is based on the average pixel value between the first pixel 501 and the third pixel 503. The other average line 526 is based on the third pixel 503 with a smaller color value. According to the second axis parallel to the vector 513, the other pixels of the photo or the pixels to which the detection area belongs are corrected. In another embodiment, the first pixel 501 with a larger color value can also be used as a reference. The example shown in Figure 5B is to correct the second axis of the photo, and also To correct the first axis.

Please refer to FIG. 5C for a schematic diagram of the two-dimensional correction slope of the color patch according to an embodiment of the present application. In addition to the slope line 520 corresponding to the second axis already shown in FIG. 5B, a slope line 530 corresponding to the first axis is also included. The slope line 530 may correspond to the vector 512 or 534 of FIG. 5A. In step 450 of FIG. 4, the slope line 530 may be corrected or equalized to an average line to correct other pixels of the photo or pixels to which the detection area belongs. The application does not limit the order of correcting the first axis and the second axis.

In step 440 shown in FIG. 4, the following judgment step may be included. That is, the two slope lines corresponding to the vectors 513 and 524 are compared, and when the slope difference between the two is too large, the color patch sample can be judged as unqualified. It is also possible to compare the two slope lines corresponding to the vectors 512 and 534 respectively. When the slope difference between the two is too large, it can be judged that the color patch sample is unqualified.

In one embodiment, the two slope lines corresponding to the vectors 513 and 524 can be averaged, and the two slope lines corresponding to the vectors 512 and 534 can be averaged. Next, the two averaged slope lines are used for correction.

In the photo 200, although the color patch sample 210 is near the defect 110, if the light source is closer to one of the two, the error caused by using the linear correction method of FIGS. 5A-5C may still be large. Therefore, in order to reduce errors, multiple color patch samples can be used for correction.

Please refer to FIG. 6, which is a schematic diagram of a photograph 600 according to an embodiment of the present application. The difference from the photo 200 is that the photo 600 includes the color patch samples 210 and the different color patch samples 620. In an embodiment, the photo 600 may include a plurality of samples of the same color patch. Such as In the embodiment of FIG. 6, the photo 600 may include a plurality of color patch samples, at least two color patch samples of which are different. Furthermore, in an embodiment, a connection line 610 of two color patch samples among the plurality of color patch samples passes through the detection area 110. The detection area 110 can be corrected by projecting the pixels of the two color patch samples 210 and 620 on the first axis and the second axis. Since the detection area 110 falls between two pixels on each axis, the detection area 110 can be better corrected.

The connection line 610 may be a connection line between the centers of the two color block samples 210 and 620, or may be a connection line of a specific point of the two color block samples 210 and 620. The connection line 610 may be a line or an area. For example, it may be a region formed by the lower left corner and the upper right corner of the color patch sample 210, and the lower left corner and the upper right corner of the color patch sample 620. The present application does not limit the shape of the color patch samples, and therefore does not limit the connecting area between the plurality of color patch samples.

In one embodiment, the color patch samples 620 may be corrected according to the color patch samples 210 alone. Since the color patch sample 620 further includes a yellow color patch 630, the characteristics of the corrected yellow color patch 630 and the color patch sample can be compared. When the error value falls within a range, it can be judged that the correction according to the color patch sample 210 is effective. When the error value exceeds a range, it can be judged that the correction based on the color patch sample 210 alone is invalid, the photo can be retaken, or the joint correction of the two color patch samples 210 and 620 can be changed instead. The present application does not limit that the color patch sample 620 can only contain a yellow color patch 630 mixed with blue and green, and can include color patches of other colors. For example, the excess color patch may be purple mixed with blue and red, or may be green-blue mixed with green and blue.

In the embodiment shown in FIG. 4, multiple pixels of a single color block sample are mainly used for correction. However, due to the size of the color block samples, the distance between these pixels is not large, and their narrow sampling space does not necessarily represent the situation of the entire photo. In the embodiment shown in Figure 6 Among them, multiple pixels of multiple color patch samples can be used for correction. Since these pixels cross over the flaws to be detected, their larger sampling space can represent the situation of the entire photo.

In the embodiment shown in FIG. 2, the direction of a single color patch sample can be aligned with the two axes of the photo through one rotation. However, in the embodiment shown in FIG. 6, the directions of the plurality of color patch samples are not necessarily aligned with each other. They are not necessarily aligned with the two axes of the photo.

Please refer to FIG. 7, which is a schematic diagram of a photo correction according to an embodiment of the present application. The photograph 700 contains two color patch samples 712 and 714. Before the whole picture can be corrected, the two color patch samples 712 and 714 are rotated respectively to generate two equal-calibrated virtual color patch samples 722 and 724, respectively, aligned with the first axis and the second axis of the photo . The equivalent calibration here means that the slopes of the color patch sample 712 and the virtual color patch sample 722 on the first axis and the second axis are equivalent, and the color patch sample 714 and the virtual color patch sample 724 are on the first axis and the first The slopes of the two axes are also comparable. The area of the color patch sample 712 and the virtual color patch sample 722 should be the same, and the area of the color patch sample 714 and the virtual color patch sample 724 should also be the same. The connection 610 of the embodiment shown in FIG. 6 can be calculated using the original two color patch samples 712 and 714, or can be calculated using the corrected two color patch samples 722 and 724 to determine whether it passes the detection area.

Please refer to FIG. 8, which is a schematic flowchart of a method 800 for epidermal layer detection according to an embodiment of the present application. Compared with the detection method 400 shown in FIG. 4, the detection method 800 is different in that a plurality of color patch samples are used. The detection method can be applied to the detection application program executed by the central processor 360 of FIG. 3, or the detection application program can be used to execute some steps of the epidermal layer detection method 800.

As mentioned above, among the steps shown in the flowcharts of this application, other steps not related to this application may be inserted. Unless there are causal dependencies, this application is also not limited The execution order of each step. The detection method 800 includes the following steps:

Step 810: Paste multiple color patch samples near the detection area of the biological surface layer. These color patch samples may be the same or different. In an embodiment, two of the plurality of color block samples pass through the area to be detected. When the organism to be detected is a cooperative target, this step can be performed using an automated machine. For example, when you want to detect a human face, you can ask the subject to be close to a face frame, and the robot arm will paste the color block sample on the human face. The face frame and the robot arm may be part of the detection device of FIG. 3, connected to the I/O device bridge 330 and controlled by the detection application executed by the central processor 360.

Step 820: Use a camera module to take a photo, the photo includes the plurality of color block samples and the detection area. The photo may be the photo 600 of FIG. 6 or the photo 700 of FIG. 7, and the camera module may be the camera module 320 of FIG. 3.

Step 830: Search and locate the plurality of color block samples in the photo according to the characteristics of the color block samples predicted in advance. The characteristics of the color patch samples referred to here may include, but are not limited to, the color, size, shape of the color patch samples and the positional relationship between them. Since there are already many open source softwares that provide the functions of face recognition or object recognition in photos, this application believes that those of ordinary skill in the art can use existing technology to search and locate the multiple color block samples in the photo, so they will not be used again. Repeat. Then, the flow may enter optional step 840 or 850, or directly enter step 860. Of course, if multiple color patch samples cannot be searched, the process returns to step 820.

Optional step 840: determine whether the plurality of color patch samples are qualified? The application may include more sub-steps to judge whether the color patch sample is qualified. For example, it can be determined whether each of the color patch samples conforms to a certain size, for example, the total area is greater than a few pixels, or the area of each color is greater than a few pixels. In another embodiment, you can judge each Among the color block samples, whether the difference between the maximum value and the minimum value of the color in each color block is less than a certain range. When the difference is too large, it may indicate that the color patch has partially faded, or is affected by the shadow and cannot be used together. In a further embodiment, it can be determined whether the maximum value of each color block in each color block sample that is not the color is greater than a certain range. For example, when the pixels of the red color block have an excessively large blue or green value, it may indicate that the red color block has partially faded, or is affected by shadows and cannot be used together.

When the judgment result of step 840 shows that at least two of the color patch samples are qualified, the flow proceeds to optional step 850 or step 860. When the judgment result shows that only one or no color patch samples are qualified, the flow returns to step 820. The above steps 820-840 can be performed in the aforementioned preview mode. When the user previews, the detection application can perform the calculation results of these steps on the current preview screen to determine whether the preview screen is qualified. When the preview screen is qualified, the user can save the photo. Otherwise, the user will not be allowed to save the photo and perform the following steps.

Optional step 850: Adjust the plurality of color patch samples according to the axial direction of the photo. When the directions of the plurality of qualified color patch samples differ from the direction of the photo by more than a range, as shown in the embodiment shown in FIG. 7, the directions of the qualified color patch samples may be adjusted into multiple virtual color patch samples respectively. The areas of these virtual color patch samples and the original qualified color patch samples should be consistent or fall within an error range, and the slope lines corresponding to the first axis and the second axis should be consistent or fall within another error range.

Step 860: Perform color correction on the photo according to the difference between the characteristics of the plurality of (virtual) color patch samples and the color patch samples. The application may include more sub-steps to perform color correction. In an embodiment, the color values of each (virtual) color patch sample can be averaged to obtain A difference in the color value characteristics of its color block samples. However, according to the position of each difference and each (virtual) color patch sample, each pixel of the photo is compensated corresponding to each difference to obtain a corrected photo. In later sections of this application, color correction will be explained further.

Step 870: Search and locate at least one detection area in the corrected photo according to the detection conditions preset in advance. The detection conditions referred to herein may include, but are not limited to, the defects, or the color, size, and shape of the detection area and the positional relationship between them. Since there are many open source softwares that provide functions for face recognition or object recognition in photos, this application believes that ordinary people in the art can use existing technology to search for and locate defects or detection areas in the photos, so they will not be repeated here. . In one embodiment, the blemishes may include, but are not limited to, red and swollen wounds, deep black scars, or surface pigmented areas. In an embodiment, the process 800 may continue to perform the optional step 875, or perform the optional step 880, 480, or 490.

Optional step 875: determine whether the detection area is between at least two (virtual) color patch samples? That is, according to the embodiment of FIG. 6, it is determined whether the connection or area between at least two (virtual) color patch samples passes through the detection area. When the connection or area does not pass the detection area, the process 800 can return to step 820 to take a new photo again. When the line or area passes the detection zone, the process 800 may continue to perform optional steps 880, 480, or 490.

Optional step 880: Calculate the detection area between the two (virtual) color patch samples based on the size recorded in the characteristics of the color patch sample and the size of the two (virtual) color patch samples in the corrected photo Size and/or area. Since the color patch sample features include the size of each color patch, the size of the detection area can be estimated based on how many pixels the color patch occupies in the photo. The area of the detection area can also be estimated based on the number of pixels occupied by the detection area. The quantified parameters such as the size and area of the defect or detection area can be used in step 480 Among them, compare with the historical records of the detection area to get the comparison result.

In an embodiment, the ratio of the size recorded in the characteristics of the plurality of color patch samples and their corresponding color patch samples may be different. In this case, the size and/or area of the detection area can be calculated using the internal difference method according to the distance between the detection area and the plurality of color patch samples. For example, when the center of the detection zone and the center of the first patch sample are the first distance D1, the center of the detection zone and the center of the second patch sample are the second distance D2, and the first patch sample is The scale of the feature is R1, and the scale of the second color block sample and its feature is R2, then the scale of the detection area can be calculated using the following internal difference method: (R1*D1+R2*D2)/(D1+D2).

Please refer to FIG. 9, which is a schematic diagram of color correction according to an embodiment of the present application. When there are more than two color patch samples on the photo, the slope line of each color of each (virtual) color patch sample on the first axis and the second axis can be calculated according to the embodiment shown in FIG. 5A. In the embodiment shown in FIG. 9, the two color patch samples have two projection positions 910 and 920 on the first axis. It has two slope lines 915 and 925 on the first axis, respectively. In the entire photo or at least in the two first-axis projection positions 910 and 920, it can be assumed that the slope is linearly between the slope lines 915 and 925. Taking the projection position 930 as an example, its distance from the center of the projection position 910 on the first axis is the first length 913, and its distance from the center of the projection position 920 on the first axis is the second length 932. Therefore, the slope line 935 of the projection position 930 can be calculated according to the internal difference algorithm of the aforementioned scale, that is, (slope 915*first length 913+slope 925*second length 932)/(first length 913+second length 932) .

Therefore, for a certain small area, each color of each color block sample can be projected on a straight line formed by the slope lines of the first axis and the second axis, respectively, to calculate the slope lines of the two axes of the small area. Then, according to the slope lines of these two axes, the pixel values in the small block area are compensated and corrected. The embodiment shown in FIG. 9 can be applied to step 860 of FIG. 8.

Please refer to FIG. 10, which is a schematic diagram of color correction according to an embodiment of the present application. When there are more than three color patch samples on the photo, the slope line of each color of each (virtual) color patch sample on the first axis and the second axis can be calculated according to the embodiment shown in FIG. 5A. The difference from the embodiment of FIG. 9 is that FIG. 10 has more than three color patch samples. An approximate quadratic curve can be obtained according to the slope values of each color projected on a certain axis on more than three. Therefore, for a small area, you can project the quadratic curve formed by the slope lines of the first axis and the second axis for each color of each color block sample, and calculate the quadratic of the two axes of the small area curve. Then, according to the quadratic curves of the two axes, the pixel values in the small block area are compensated and corrected. In another embodiment, in the small area, a slope line can be used instead of the quadratic curve in order to reduce the amount of calculation. The embodiment shown in FIG. 10 can also be applied to step 860 of FIG. 8.

According to an embodiment of the present application, a biological epidermal layer detection method is provided, which is used to provide a quantitative parameter of a detection area, including: using a camera module to take a photo, wherein the photo includes one or more color patch samples and a biological epidermis The detection area on the layer; searching and locating one or more of the color patch samples in the photo according to the pre-predicted feature values of the plurality of color patch samples; according to one or more of the color patch samples and the plurality of color patches Correct the photo according to the difference of the sample characteristic value; search and locate the detection area in the corrected photo according to the detection conditions preset in advance; and calculate the quantization parameter of the detection area.

In one example, in order to speed up the working time and reduce the operating burden, the biological epidermal layer detection method further includes: before taking the photo, a mechanical device is used to fix the biological epidermal layer to be detected, and one or more samples of the color patch Stick to the biological epidermis.

In an example, in order to ensure the reliability and correctness of the calibration, the biological epidermal layer detection method further includes: judging whether each color patch sample is qualified, when there is at least one color patch sample When the book is qualified, the steps of correcting the photo are carried out. In an example, the step of judging whether each color patch sample is qualified further includes one of the following steps or any combination thereof: judging whether the color patch sample is larger than a certain size, and when smaller than the size, the color patch The sample is judged to be unqualified; the total area of the color patch sample is judged to be greater than a certain number of pixels, and when it is less than the number of pixels, the color patch sample is judged to be unqualified; each color of the color patch sample is judged Whether the difference between the maximum value and the minimum value of the color in the block is less than a certain range, when the difference is greater than the range, the color block sample is judged as unqualified; and each color of the color block sample is judged Whether the maximum value in the color block that is not the color is greater than a certain range, and when the maximum value is greater than the range, the color block sample is judged as unqualified.

In an example, in order to provide a calibration step for interactively verifying different color block samples, when there are more than two qualified color block samples, at least two qualified color block samples are different, and the two different colors The block sample includes a first color block sample and a second color block sample. The second color block sample includes a color block that is not included in the first color block sample. In an example, the bioepidermal layer detection method further includes: correcting the second color patch sample based on the difference between the characteristic values of the first color patch sample and the plurality of color patch samples, when the color patch and its characteristic value If the error is less than a range, then correct the photo.

In an example, in order to ensure that the internal difference method used for calibration is applicable, when more than two samples of the color patch are qualified, the detection method further includes determining whether there are two A connection line of the color patch sample passes through the detection area. When the connection line does not pass the detection area, the quantization parameter of the detection area is not calculated. In an example, the two ends of the connection are the centers of the two color patch samples. In an example, the two ends of the connection are respectively arbitrary pixels of the two color patch samples.

In an example, in order to use a single color block sample for correction, when searching and locating a color block sample, the above-mentioned step of correcting the photo further includes: obtaining two of the color blocks of the color block sample Pixels; calculating a slope line corresponding to the color patch among the two pixels; and correcting and compensating the remaining pixels of the photo according to the slope line and the color. In an example, for convenience of calculation, the connection line of the two pixels is parallel to a first axis or a second axis of the photo, wherein the first axis is perpendicular to the second axis. In an example, in order to obtain the longest distance, the two pixels are located at the corners or edges of the color patch, respectively.

In an example, in order to use the inner-difference method of two color patch samples, when searching and locating two color patch samples, the above-mentioned step of correcting the photo further includes: obtaining two of the color patch samples separately One pixel of each color patch; calculate a slope line corresponding to the color patch among the two pixels; and correct and compensate the remaining pixels of the photo according to the slope line and the color .

In an example, in order to use the inner-difference method of two color patch samples, when searching for and locating a first color patch sample and a second color patch sample, the above step of correcting the photo further includes: obtaining Two first pixels of one color patch of the first color patch sample; calculating a first slope line corresponding to the color patch of the two first pixel dots; obtaining the second color patch sample Where two second pixels of the color patch; among the two second pixels, calculate a second slope line corresponding to the color patch; for a pixel area of the photo, according to the distance of the pixel area Calculating a slope line for a first length of the first color patch sample, a second length of the pixel area from the second color patch sample, the first slope line and the second slope line; and according to The slope line and the color correct and compensate the pixel area, where the two first The pixel line, the two second pixel points, the first length and the second length are all parallel to the same direction.

In an example, in order to use the internal difference method of three color patch samples, when searching for and locating a first color patch sample, a second color patch sample, and a third color patch sample, the above is performed on the photo The calibration step further includes: obtaining two first pixels of one color patch of the first color patch sample; calculating a first slope line corresponding to the color patch among the two first pixel dots; Obtaining two second pixels of the color block in the second color block sample; calculating a second slope line corresponding to the color block among the two second pixels; obtaining the third color block The two third pixels of the color patch in the sample; calculating a third slope line corresponding to the color patch among the two third pixels; according to the first slope line and the second slope The degree line and the third slope line and the three positions projected on an axis to calculate the quadratic curve of a color block sample; for a pixel area of the photo, according to the pixel area A first length, a second length of the pixel area from the second color block sample, a third length of the pixel area from the second color block sample and a quadratic curve of the color block sample, calculate a quadratic curve A curve; and correcting and compensating the pixel area according to the quadratic curve and the color, wherein the connection of the two first pixels, the connection of the two second pixels, and the two third pixels , The first length, the second length and the third length are all parallel to the same direction.

In an example, in order to output the quantization parameter, when searching and locating a color block sample, the above step of calculating the quantization parameter of the detection area further includes: according to the size and the color block recorded in the characteristic value of the color block sample The size of the sample in the corrected photo is used to calculate the size and/or area of the detection area.

In an example, in order to output quantization parameters, when searching and locating a first color When a block sample and a second color block sample, the above step of calculating the quantization parameter of the detection area further includes: according to the size recorded in the characteristic value of the color block sample and the first color block sample in the corrected photo Calculate a first scale; calculate a second scale based on the size recorded in the feature value of the color patch sample and the size of the second color patch sample in the corrected photo; based on the detection area and the first color A first distance of the block sample, a second distance of the second color block sample in the detection area, the first scale and the second scale, a detection area scale is calculated; according to the detection area scale and the detection area in the After correcting the size in the photo, calculate the size and/or area of the detection area.

In an example, in order to output quantitative parameters, the biological epidermal layer detection method further includes calculating one or any combination of the following parameters: the size of the detection area; the area of the detection area; the center of the detection area; The center of gravity of the detection area with weights; the average color of the detection area; the maximum value of a color in the detection area; and the minimum value of a color in the detection area.

In an example, in order to compare with the historical record, the biological epidermal layer detection method further includes: reading the corresponding historical parameter of the detection area from a memory module; and calculating the comparison result of the historical parameter and the parameter.

According to an embodiment of the present application, a biological epidermal layer detection device is provided for providing quantization parameters of a detection area, including: a camera module for taking a photo, wherein the photo includes one or more color patch samples And the detection area on the biological epidermis; a memory module for storing a detection application, the photo and the characteristic values of a plurality of color patch samples; and a central processor connected to the camera module and the memory The body module is used to execute the following steps of the detection application: search and locate one or more of the color patch samples in the photo according to the previously predicted feature values of the plurality of color patch samples; according to one or more of the Color block samples and the plurality of color block samples The difference in the feature value corrects the photo; searches and locates the detection area in the corrected photo according to the detection conditions preset in advance; and calculates the quantization parameter of the detection area.

According to an embodiment of the present application, a biological epidermal layer detection system is provided for providing a quantitative parameter of a detection area, including: at least one color patch sample; and a biological epidermal layer detection device. The embodiment of the biological epidermal layer detection device is as described above.

According to the above embodiments, the present application provides a method, device, and system that can quantitatively present the results of epidermal layer detection, and can accurately evaluate the effects of drugs or maintenance products.

800‧‧‧Epidermal layer detection method

810~880‧‧‧Step

Claims (3)

A biological epidermal layer detection method for providing quantitative parameters of a detection area, including: using a camera module to take a photo, wherein the photo includes one or more color patch samples and the detection area on the biological epidermal layer; The feature values of the plurality of color patch samples predicted in advance, search and locate one or more of the color patch samples in the photo; according to the difference between the feature values of the one or more color patch samples and the plurality of color patch samples, Correct the photo; search and locate the detection area in the corrected photo according to the detection conditions preset in advance; and calculate the quantization parameters of the detection area, where when two color patch samples are searched and located, the above The step of correcting the photo further includes: obtaining one pixel of each color patch of the two color patch samples; calculating a slope line corresponding to the color patch among the two pixel dots; and according to the The slope line and the color correct and compensate the remaining pixels of the photo. A biological epidermal layer detection device for providing quantitative parameters of a detection area, including: a camera module for taking a photo, wherein the photo includes one or more color block samples and the detection area on the biological epidermal layer A memory module for storing an inspection application, the photos and the characteristic values of a plurality of color samples; and a central processor connected to the camera module and the memory module for performing the inspection should Use the following steps of the program: search and locate one or more of the color block samples in the photo according to the pre-predicted feature values of the plurality of color block samples; according to one or more of the color block samples and the plurality of color blocks The difference in the sample characteristic value, correct the photo, search and locate the detection area in the corrected photo according to the preset detection conditions; and calculate the quantitative parameters of the detection area, where two colors are searched and located In the case of a block sample, the above-mentioned step of correcting the photo further includes: obtaining one pixel of each color block of the two color block samples; and calculating the corresponding value of the color block among the two pixels A slope line; and according to the slope line and the color, correct and compensate the remaining pixels of the photo. A biological epidermal layer detection system for providing quantitative parameters of a detection area, including: at least one color block sample; and a biological epidermal layer detection device, including: a camera module for taking a photo, wherein the photo contains One or more of the color patch samples and the detection area on the biological epidermis; a memory module for storing a detection application, the photo and the characteristic values of the plurality of color patch samples; and a central processing unit, connected To the camera module and the memory module for performing the inspection Measure the following steps of the application: search and locate one or more of the color block samples in the photo according to the pre-predicted feature values of the plurality of color block samples; based on the one or more color block samples and the multiple colors Correct the photo based on the difference in the feature values of the block samples; search and locate the detection area in the corrected photo according to the detection conditions preset in advance; and calculate the quantization parameters of the detection area, when searching and locating two In the case of color patch samples, the above step of correcting the photo further includes: obtaining one pixel of each color patch of the two color patch samples; calculating the two pixels corresponding to the color patch A slope line of; and according to the slope line and the color, correct and compensate the remaining pixels of the photo.

Images