JP4116709B2 - Omnidirectional 1D barcode reader - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an omnidirectional one-dimensional bar code reading apparatus that uses a CCD camera to read and identify a bar code label indicating a product type affixed to a product having a different shape moving on a belt conveyor in a factory.
[0002]
[Prior art]
Conventionally, to read the bar code label attached to different product shapes moving on the belt conveyor, for example, using a bar code reader which uses a record over Heather light. At that time, in order to increase the reading accuracy, the position where the bar code is pasted, the direction of the bar code, how to put it on the belt conveyor, etc. are devised. Further, every time the shape (size) of the product with the barcode label attached is changed, the reader itself is moved up and down to keep the distance between the product and the reader constant.
[0003]
[Problems to be solved by the invention]
As described above, in order to attach a barcode label to a product moving on a belt conveyor in a factory and identify the product, it is necessary to read the barcode attached for product identification. At that time, the product has a different shape (size), it is uncertain where it will be placed on the conveyor, and the barcode label is not defined in which direction on the product surface. Bar code must be read under.
[0004]
In the case of a conventional barcode reader, for example, a barcode reader using laser light, it is necessary to make the position and direction of attaching the barcode, and how to place the barcode on the belt conveyor constant. Further, the conventional barcode reader has a shallow reading depth, and reading becomes impossible when the distance from the barcode exceeds a certain value. Therefore, in order to read and identify barcode labels attached to products of different shapes moving on the belt conveyor, the reader must be moved in accordance with the shape (size) of the product. Further, since the conventional barcode reading apparatus has a constant scanning direction at the time of reading, it may not be read if the barcode direction is indefinite.
[0005]
The present invention has been made to solve the above-described problems. Even when the shape of a product with a barcode label is different, barcode reading and identification can be performed without moving the barcode reading device. An object of the present invention is to provide an omnidirectional one-dimensional bar code reader that can identify all directions even when the direction of the bar code is indefinite. It is.
[0006]
[Means for Solving the Problems]
To achieve the above object, the present invention uses a CCD camera to capture an image of the entire barcode label as a subject, and performs image processing based on the captured image to enable barcode reading and identification. It is characterized by that. That is, the present invention provides the following means (1) and (2) .
[0007]
(1) imaging a bar code label attached to the product as more image on the CCD camera, a bar code reader of an image obtained by image processing reads the bar code, the position of the bar code label A first CCD camera for identifying the image, a second CCD camera for capturing the barcode identified by the first CCD camera as an input image, and an image obtained by imaging with the CCD camera. to configure the system with an image processing operation unit for performing processing.
[0008]
Upper Symbol image processing calculation unit a predetermined level threshold to the luminance value with based on the bar code image from the CCD camera for capturing a bar code as an input, performs the projection against the cut out portion of the bar code The binarization process is performed by setting the value to a value that can recognize the bar and space part based on the projection view, and that the change of the part corresponding to the bar and space does not vary significantly by changing the threshold value. When the barcode is identified and cannot be identified only by the binarization process, the luminance of the difference from the maximum luminance value at each pixel in the area of each bar in the projection view is obtained and summed Is determined as the luminance area, and the ratio is compared for each bar area to determine the narrow bar and the wide bar, and the processing for comparing the area ratio of the luminance similar to the above processing is also performed for the space. Narro Determine the space and wide space, and add the characteristics peculiar to each bar code to the above determination result. If you cannot identify the bar code, reset the threshold and repeat the above operation. identifiable and readable bar codes that have an identification algorithm for faster than distant to all directions one-dimensional bar code reader.
[0009]
( 2 ) The image processing calculation unit of the barcode reading apparatus sets the threshold value to the image input from the CCD camera and compares the luminance area ratio even when the direction of the barcode label is unspecified. The omnidirectional one-dimensional barcode reading apparatus according to ( 1 ), wherein the barcode can be identified.
[0010]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0011]
The present invention was originally developed for the purpose of non-contact reading and identification of barcodes attached to products having different shapes that move at high speed on a belt conveyor in a factory. FIG. 1 shows a barcode photographing environment targeted by the present invention, in which the moving speed υ of the belt conveyor 1 is set to 15 to 20 m / min, and the distance between the barcode 3 and the CCD camera 2 is set to 2 m. .
[0012]
FIG. 2 is a configuration diagram of an omnidirectional one-dimensional barcode reading apparatus which is a system for automatically identifying a barcode according to an embodiment of the present invention. In the figure, first, the entire surface of the object moving on the belt conveyor is imaged by the # 1 CCD camera 2-1. Next, the image processing unit 4 and the host computer 5 perform image processing on the image obtained by the above imaging to detect the position of the barcode label. The image processing unit 4 mainly performs image processing such as binarization and recognition, and the host computer 5 performs control of the timing of the entire system, monitoring, debugging, and the like.
[0013]
This result is transmitted from the host computer 5 to the # 2 CCD camera 2-2, and the # 2 CCD camera 2-2 narrows the object to be photographed to a barcode, and the barcode portion is input to the image processing unit 4 as an input image. is there. The image obtained by the # 2 CCD camera 2-2 is subjected to image division processing by the image processing unit 4 and the host computer 5, whereby the barcode shape can be extracted and identified.
[0014]
The shape extraction process for cutting out the barcode portion will be described below based on the image captured by the # 2 CCD camera 2-2. FIG. 3 shows the barcode label photographing conditions of the # 2 CCD camera 2-2 in FIG. 2 and the specifications of the camera used in the experiment. In the figure, 7 and 8 indicate illumination, 9a is an image range when the distance is 1 m, and 9b is an image range when the distance is 2 m. FIG. 4 shows an image of a barcode label photographed under the conditions of FIG.
[0015]
The shooting conditions were a lens of 90 [mm], an aperture of 5.0, and an illuminance of 1000 [lux]. The equipment used is a 3 CCD camera (DXC-950 SONY), the number of effective pixels is 768 x 494, the horizontal resolution is 750 TV lines, the illuminance is 2000 lx (F9.5, 3200K), the zoom lens (VCL-712BX SONY) and the focal length is 7.5 to 90 mm, and the aperture is 1.4 to 16.
[0016]
4A and 4B are diagrams showing an image of a bar code label photographed under the above conditions. FIG. 4A shows a case where the distance between the CCD camera and the bar code is 1 m, and FIG. 4B shows a case where the distance is 2 m. FIG. 5 shows the result of edge detection and binarization for extracting the barcode shape from the original image of FIG.
[0017]
The threshold value 200 is an 8-bit A / D converter (analog / digital converter) converted into a digital signal when the video signal from the camera is converted to a digital signal, so the video is represented by a digital value of an integer from 0 to 255. The luminance value of the signal is displayed. Therefore, it means that 200 of these 0-255 is selected. As shown in this figure, the bar portion is somewhat missing, but sufficient results were obtained for shape extraction.
[0018]
Therefore, next, we examined shape extraction when the distance between the barcode and the CCD camera was 1 m or more. The bar code is composed of a black bar and a white space, and each bar and space has a narrow width and a wide width. Usually, the ratio of narrow to wide is 1: 2 to 1: 3. In identifying the barcode, it is important to accurately determine the narrow and wide.
[0019]
Next, processing is performed on an image when the distance between the barcode and the CCD camera is 1.5 m. As a processing method, binarization that is effective when the distance is 1 m is used. FIG. 6A is a diagram showing a part of the original image captured when the distance between the barcode and the CCD camera is 1.5 m, and FIG. 6B is projected onto the result of the binarization process. The threshold is set to 200.
[0020]
Comparing the result of binarization processing with the original image from FIG. 6 (b), a result capable of sufficiently extracting the shape was obtained. This indicates that binarization is effective for shape extraction even when the distance between the barcode and the CCD camera is 1.5 m. By performing binarization processing to extract the shape of the barcode, it is possible to cut out the barcode portion from the original image.
[0021]
Next, a result of studying a method for identifying a barcode cut out from an original image will be described. First, the cut image was projected. FIG. 7 is a diagram showing the result, FIG. 7 (a) is a diagram showing the projection result, FIG. 7 (b) is a result of the binarization processing, and a threshold value as shown in FIG. 7 (a). (X) is provided to perform binary conversion of the image. From FIG. 7B, it is determined that 1 or 2 pixels are narrow and 3 or 4 pixels are wide with respect to the bar and space.
[0022]
FIG. 8 shows the relationship between the barcode, the distance of the CCD camera, and the threshold value when this determination method is used. It can be seen from FIG. 8 that the identification rate depends on how the threshold is selected. The identification rate is high in the case of 70% to 80%. However, when the distance between the barcode and the CCD camera is 1.7 m, it cannot be identified even if the threshold value is changed. In the identification method using the binarization process and the pixel width, the distance limit is 1.6 m. However, it can be seen that the object of the present invention is that the distance between the barcode and the CCD camera is 2 m, so that the method using the pixel width cannot be identified.
[0023]
In view of this, a method for discriminating even when the distance between the barcode and the CCD camera is 1.7 m or more was further examined. FIGS. 9A and 9B are diagrams showing this method. FIG. 9A is a flowchart of the identification step, FIG. 9B is a diagram when the method of FIG. 9A is performed, and the distance between the barcode and the CCD camera is 2 m. FIG. Here, WB and NB in FIG. 9 represent actual bar states.
[0024]
For the projection of FIG. 9B, the threshold is set to 80% from the result of FIG. At this time, one pixel can be identified as a narrow bar, but two pixels cannot be identified. Therefore, paying attention to the shaded portion in FIG. 9, the shaded portion (1) and the shaded portion (2) are 3780 and 3789 by multiplying the luminance value and the bar width. The shaded area (3) was 1850. Since these area ratios are approximately 2: 2: 1, it is determined that the shaded portion (1) and the shaded portion (2) are wide bars and the shaded portion (3) is a narrow bar. Further, the narrow space NS and the wide space WS can be similarly identified by using the area.
[0025]
Next, the algorithm will be described. First, a barcode is cut out and projected, and binarization is performed on the projection using a threshold value. In this case, the threshold calculating a value that does not vary significantly by projection drawing can recognize the portion of the bars and spaces based on, and changes in the width of the portion corresponding to the bar and space changes the threshold To do. Next, using the result of binarization, the sum of values obtained by subtracting the luminance value from the maximum luminance value (255 × 5) of some pixels corresponding to each bar is calculated, and Wide bar and narrow bar are judged by numerical value. The same processing is performed for the space. The final determination is performed by adding a characteristic peculiar to each barcode to the result. If the identification is impossible, the process returns to the binarization process based on the threshold value after cutting out the barcode, and if the identification is still impossible, the process returns to the first barcode extraction and projection process.
[0026]
Furthermore, the type of barcode used as a target here is Code-39, and the basic configuration of this code is narrow space NS, The identification rate can be improved by determining the wide space WS. Using the above identification method, it was confirmed that the barcode can be identified even when the distance between the barcode and the CCD camera is 2 m.
[0027]
In the processes shown so far, the direction of the barcode is constant. Therefore, next, the barcode CCD camera distance was set to 2 m, and the identification when the barcode direction was different was also examined. 10A and 10B are diagrams illustrating an original image and a projection result when the barcode has an angle with respect to the projection direction. FIG. 10A is a diagram illustrating the original image when the rotation angle of the barcode is 10 degrees. These are projection views when the rotation angle of the barcode is 10 degrees. Here, as shown in FIG. 10A, the projection is processed with a narrow field. Further, from FIG. 10B, the area ratio between the narrow space and the wide space is about 1: 3. Therefore, when the angle of the barcode is small, the identification can be performed without changing the projection direction. I understand that.
[0028]
Next, the identification when the bar code angle was larger than the above was examined. Here, since the apparatus used in the present invention cannot project in an oblique direction, the image is rotated by affine exchange and projected. FIG. 11 is a projection view of the result of projection on the original image at that time, (a) is a projection view when the rotation angle of the barcode is 22.5 degrees, and (b) is a rotation of the barcode. It is a projection view when the angle is 45 degrees. In this projection result, the brightness value of the space portion varies and the accumulation of the brightness value due to the projection does not become constant, so the maximum value of the brightness value is 80%.
[0029]
From FIG. 11, it was confirmed that the area ratios of the bar portions were both about 1: 7, so that sufficient identification could be performed. From the above method, identification is possible even when the barcode direction is slightly inclined. By changing the scanning angle with respect to the image projected by this principle, omnidirectional reading becomes possible.
[0030]
【The invention's effect】
As described above, the present invention has the basic structure of the invention described in claim 1, and according to the present invention, the following effects can be obtained.
[0031]
When reading the barcode label attached to the product, there is no need to devise where on the product surface and in what direction the barcode label is attached, and the position of the product on the belt conveyor may be unspecified. .
[0032]
In addition, when reading a barcode, there is no need to reduce or stop the belt conveyor transfer speed, and the production capacity of the production line is improved.
[0033]
Furthermore, for all products transferred on the belt conveyor, the bar code label affixed to the surface of the product can be reliably read without any information such as product dimensions in advance. The type (size) of the product transferred on the line can be changed, and the production plan can be easily changed.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a barcode imaging environment of an omnidirectional one-dimensional barcode reader according to an embodiment of the present invention.
FIG. 2 is a system configuration diagram of an omnidirectional one-dimensional barcode reading apparatus according to an embodiment of the present invention.
FIG. 3 is a view showing bar code label photographing conditions of the omnidirectional one-dimensional bar code reader according to the embodiment of the present invention.
FIG. 4 is an original image of a barcode imaged by a CCD camera of an omnidirectional one-dimensional barcode reader according to an embodiment of the present invention, where (a) is a distance between the barcode and the CCD camera of 1 m. (B) is a figure in case the distance is 2 m.
FIG. 5 is a CCD camera of an omnidirectional one-dimensional barcode reader according to an embodiment of the present invention, and a barcode imaged at a distance of 1 m between the barcode and the CCD camera is subjected to shape extraction and barcode partial segmentation. It is the figure which carried out the binarization process for this.
FIG. 6 shows a barcode imaged at a distance of 1.5 m between the barcode and the CCD camera by the CCD camera of the omnidirectional one-dimensional barcode reading apparatus according to the embodiment of the present invention. The figure which shows an image, (b) is the figure which projected the result which binarized the original image.
FIG. 7 shows a barcode imaged at a distance of 1.5 m between the barcode and the CCD camera by the CCD camera of the omnidirectional one-dimensional barcode reader according to the embodiment of the present invention. The figure which shows the result of having cut out barcode based on the image, and having projected, (b) is the figure which provided the threshold value based on the projection figure, and performed the binarization process.
FIG. 8 is a diagram showing a barcode identification rate when the threshold value is changed by the omnidirectional one-dimensional barcode reader according to the embodiment of the present invention.
FIG. 9 is a diagram in which projection is performed based on an image obtained by capturing a barcode at a distance of 2 m between the barcode and the CCD camera by the omnidirectional one-dimensional barcode reading apparatus according to the embodiment of the present invention; a) is the processing flow, and (b) is a projection view.
FIG. 10 is an omnidirectional one-dimensional barcode reading apparatus according to an embodiment of the present invention, and shows a barcode imaged when the distance between the barcode and the CCD camera is 2 m and the barcode rotation angle is 10 degrees. (A) is an original image, (b) is a projection view.
FIG. 11 is a diagram of an omnidirectional one-dimensional barcode reading apparatus according to an embodiment of the present invention projected based on an original barcode image captured at a distance of 2 m between the barcode and the CCD camera; (a) is a diagram when the rotation angle of the barcode is 22.5 degrees, and (b) is a diagram when the rotation angle is 45 degrees.
[Explanation of symbols]
1 Belt conveyor 2 CCD camera 2-1 CCD camera for position detection 2-2 CCD camera for reading 3 Bar code 4 Image processor 5 Host computer 6 Product

Claims (2)

Imaging a bar code label attached to the product as more image on the CCD camera, the image obtained by image processing a bar code reading device for reading the bar code, identifies the position of the bar code label a first CCD camera for a second CCD camera for capturing a bar code identified by the same first CCD camera as an input image, the image processing to the image obtained by imaging by the CCD camera The image processing operation unit constitutes a system, and the image processing operation unit projects a barcode image from a barcode image from a CCD camera for capturing the barcode as an input. The bar and space part can be recognized on the basis of the projection drawing and the threshold value of the predetermined level for the luminance value can be recognized. If the change in the width of the bar code is set to a value that does not vary significantly by changing the threshold value, binarization processing is performed to identify the barcode, and if the binarization processing alone cannot identify The brightness of the difference from the maximum brightness value at each pixel in each bar area in the projected view is calculated, the total value is calculated as the brightness area, and the ratio is compared for each bar area, so that the narrow bar and the wide bar Bars are determined, and the same area ratio of brightness is compared with the above processing for the space to determine narrow space and wide space. adding to, if not be identified, re-setting the threshold, an identification algorithm that allows identifying the bar code by repeating the above operation, and capable of reading bar codes faster than distant DOO omnidirectional one-dimensional bar code reading apparatus according to claim. Image processing calculation unit of the upper Symbol bar code reader, by comparing the area ratio of the luminance with the direction of the bar code label even if unspecified, setting the threshold to the image input from the CCD camera omnidirectional one-dimensional bar code reading apparatus according to claim 1, identifiable barcodes by.

Images