JP2014053859A - Mobile object observation system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To efficiently store image data without reducing an information amount of regions of a mobile object and a desired still object while capacity of the image data is compressed.SOLUTION: The mobile object observation device includes: an imaging apparatus (10) for imaging a plurality of continuous frames; a mobile object detecting section (112) for detecting a mobile object in response to the plurality of imaged frames; a still object detecting section (113) for detecting a still object in response to the imaged frame; and a partial image generating section (114) for cutting out respective regions of the detected mobile object and still object. The still object detecting section detects the still object in at least one frame including the mobile object in the frames from the frame where the mobile object detecting section detects the mobile object to the frame where the mobile object is eliminated.

Description

  The present invention relates to a moving object observation apparatus.

  In the surveillance camera field, as a technique for compressing image data, there is known an apparatus that detects a moving object and converts and outputs an image with a different number of gradations in a moving object region and a region other than the moving object (Patent Document 1). ).

JP 2009-253747 A

  By the way, in astronomical observation, when moving objects such as meteors and space debris are observed with moving images, not only information on moving objects but also information on the position and brightness of a star observed as a stationary object is important. This is because the relative brightness of meteors and space debris can be calculated on the basis of the brightness of known stars, and the distance and size of the moving object from the ground can be calculated from the information. Therefore, in addition to the image information of the moving object, it is necessary to store the luminance information of the star observed as at least one stationary object without being compressed. However, since the gradation other than the moving object detection area is compressed, there is a problem that the gradation information of the stationary object as a marker is reduced.

  An object of the present invention is to efficiently store image data while reducing the amount of information of the area of the moving object and a desired stationary object while compressing the capacity of the image data.

  The moving body observation apparatus of the present invention includes an imaging apparatus that captures a plurality of consecutive frames, a moving body detection unit that detects a moving body based on the plurality of captured frames, and the captured frames. A stationary object detection unit that detects a stationary object, and a partial image generation unit that cuts out the areas of the detected moving object and the stationary object, respectively, and the stationary object detection unit is moved by the moving object detection unit. The stationary object in at least one frame including the moving body is detected from the frame in which the moving body is detected to the frame in which the moving body disappears.

  The image data can be efficiently stored without reducing the amount of information of the area of the moving object and the desired stationary object while compressing the capacity of the image data.

It is a figure which shows the structural example of a moving body detection apparatus. It is a flowchart which cuts out a moving body and a stationary object. It is a schematic diagram of the cut-out area | region of a mobile body detection and a stationary object detection. It is a figure which shows an example of the positional information and luminance information addition of a moving body and a stationary object. It is a flowchart in the case of cutting out and storing a stationary object. It is a schematic diagram of the cut-out area | region of a mobile body detection and a stationary object detection. It is the schematic diagram which detected the mobile body estimation image area | region of the flame | frame before and behind a mobile body detection. It is a flowchart in the case of cutting out and preserve | saving. It is the schematic diagram which detected the mobile body estimation image area | region of the flame | frame before a mobile body detection. It is the figure which showed the position coordinate and luminance information of the flame | frame before and behind a mobile body detection. It is the figure which showed the position coordinate of the flame | frame before and behind a mobile body detection.

(First embodiment)
FIG. 1 is a diagram illustrating an example of a configuration of a moving object observation apparatus according to the first embodiment of the present invention. The moving body observation device 1 includes an imaging device 10, an image processing device 11, and an image storage device 12. The image processing apparatus 11 includes an image input unit 110, a temporary storage unit 111, a moving body detection unit 112, a stationary object detection unit 113, a partial image generation unit 114, a control unit 115, and an image output unit 116. The image pickup apparatus 10 includes an optical system and an image pickup element. The image of the subject is formed on the image pickup element by the optical system, and the subject light is converted into an electric signal (image signal) by the image pickup element. For example, a lens or an astronomical telescope can be used as the optical system, and a CCD image sensor or a CMOS image sensor can be used as the imaging device. The imaging device 10 captures and outputs images of a plurality of consecutive frames periodically at a predetermined imaging time (for example, 16.6 milliseconds). The image input unit 110 performs various signal processing on the electrical signal (image signal) converted by the imaging device 10. For example, when the electrical signal is an analog signal, the image input unit 110 converts the analog signal into a digital signal. The temporary storage unit 111 temporarily stores an input signal from the image input unit 110 for each image frame. The temporary storage unit 111 can temporarily store several tens to several hundred frames. When data exceeding the capacity of the temporary storage unit 111 is input, the data is overwritten in order from the past data. For example, the temporary storage unit 111 can use a semiconductor memory such as a DRAM (dynamic random access memory).

  The moving body detection unit 112 processes the image signal stored in the temporary storage unit 111 and detects the moving body. That is, the moving body detection unit 112 detects a moving body based on a plurality of frames imaged by the imaging device 10. The algorithm for detecting the moving object is not particularly limited. For example, a method of taking a difference between two frame images and detecting the moving object when the difference value exceeds a threshold value can be used. The moving body detection unit 112 outputs the result of detecting the moving body to the control unit 115. For example, the moving body detection unit 112 can be used by being implemented as a functional block in an FPGA (Field-Programmable Gate Array). The stationary object detection unit 113 processes the image signal stored in the temporary storage unit 111 and detects a stationary object. That is, the stationary object detection unit 113 detects a stationary object based on the frame imaged by the imaging device 10. The present embodiment is characterized in that the stationary object detection unit 113 performs stationary object detection only on an image in which the moving body detection unit 112 has detected the moving body. The stationary object detection unit 113 performs object extraction in an image obtained by masking and removing the region including the moving object. As a method for extracting a stationary object, for example, there is a method in which an image is binarized with a certain threshold value and an object having a certain size and brightness is extracted. Further, the object to be extracted may be set by the user. For example, the stationary object detection unit 113 can be used by being implemented as a functional block in an FPGA (Field-Programmable Gate Array).

  The partial image generation unit 114 determines the area of the moving object detected by the moving object detection unit 112 and the area of the stationary object detected by the stationary object detection unit 113 based on the plurality of frames captured by the imaging device 10. cut. The partial image generation unit 114 inputs the position information of the moving body from the moving body detection unit 112 and inputs an image from the temporary storage unit 111 when the moving body is detected. Then, the partial image generation unit 114 cuts out a peripheral area including the moving body from the entire image detected by the moving body. In addition, the partial image generation unit 114 receives the position information of the stationary object from the stationary object detection unit 113, and cuts out a peripheral region including the stationary object. Details will be described later. For example, the partial image generation unit 114 can be used by being implemented as a functional block in an FPGA (Field-Programmable Gate Array).

  The control unit 115 receives a detection signal from the moving body detection unit 112 and an input signal from a user setting unit (not shown), and outputs which image among the partial images input to the image output unit 116 to the image storage device 12. Control what to do. When the moving body is detected, the control unit 115 controls to output a cut-out image around the moving body. In addition, the control unit 115 performs control by determining whether to output a cutout image of a stationary object according to the number of frames after the moving object is detected. For example, the control unit 115 can be used by being implemented as a functional block in an FPGA (Field-Programmable Gate Array).

  The image output unit 116 outputs the image information generated by the partial image generation unit 114 to the image storage device 12 based on the control of the control unit 115. For example, the image output unit 116 can be used by being implemented as a functional block in an FPGA (Field-Programmable Gate Array). The image storage device 12 stores the image information and other information output from the image output unit 116. The image storage device 12 may be non-volatile, and a hard disk (HDD), a flash memory drive (SSD), or the like can be used. The image storage device 12 may be a server on a network and connected to the image processing device 11 via a network.

  Next, details of the operation from the moving object detection to the image output will be described with reference to FIGS. A procedure of image processing in an image group from detection of a moving object to disappearance of the moving object will be described with reference to FIGS. A description will be given by taking as an example an image of the Nth frame in which the moving body 30 in FIG. 3 is detected. FIG. 2 shows a flow of processing a certain frame image. In step S <b> 20, the image input unit 110 acquires image data from the imaging device 10 and stores it in the temporary storage unit 111. Next, in step S <b> 21, the moving body detection unit 112 uses the N−1 frame image data stored in the temporary storage unit 111 to obtain a difference between the N−1 frame and N frame images. Thereby, the mobile body detection part 112 detects whether a mobile body exists. When the moving body is not detected, the subsequent image calculation is not performed. If a moving object is detected, in step S22, the moving object detection unit 112 outputs the position information of the moving object to the partial image generation unit 114, and the partial image generation unit 114 inputs an image from the temporary storage unit 111. Then, a process of cutting out the area including the moving body is performed. At this time, when a plurality of moving bodies are detected, the moving body detecting unit 112 detects all of the plurality of moving bodies, and the partial image generating unit 114 performs a process of cutting out an area including the plurality of moving bodies. At this time, the region including the moving body is a region where pixels are continuous. Although the area including the moving body 30 in FIG. 3 is shown as a rectangle, the cutout shape is not limited to a rectangle, and it is sufficient that the image data is continuous in both the horizontal and vertical directions. Image data including the cut out moving body 30 is temporarily stored in a buffer in the image output unit 116.

  In step S23, the stationary object detection unit 113 detects a stationary object. The stationary object detection unit 113 performs stationary object extraction on an image obtained by masking and excluding the region including the moving body 30 described above. As a method for extracting a stationary object, for example, there is a method in which an image is binarized with a certain threshold value and an object having a certain size and brightness is extracted. The object to be extracted may be set by the user. By detecting at least one stationary object with known brightness, the luminance of the stationary object can be compared with the luminance data of the moving object. When the stationary object is a known object and the luminance is known, the luminance of the stationary object can be calculated by relative comparison. In order to calculate the absolute position of the moving object with high accuracy, it is more desirable to detect two or more stationary objects. When calculating the absolute position of the moving object on the celestial sphere, it is calculated from the direction of the imaging device 10 and the relative position of the stationary object. When only one point is extracted as a stationary object, the deviation between the setting of the orientation of the imaging device 10 and the actual one affects the absolute position accuracy of the moving body 30. For example, when the imaging apparatus 10 is not parallel and slanted as expected, an error occurs in the position of the moving body 30 on the celestial sphere due to the inclination of the imaging apparatus 10. On the other hand, when two or more points are detected as stationary objects, the position of the moving object can be calculated from the positions of two stationary objects whose positions on the celestial sphere are already known. Can be removed.

  Next, a process of cutting out the peripheral area of the stationary object is performed. The stationary object detection unit 113 outputs the detected position information of the stationary object to the partial image generation unit 114. Similar to the moving object clipping process, the stationary object clipping area is a continuous area without thinning. Although the area including the stationary objects 31 and 32 in FIG. 3 is indicated by a rectangle, the cutout shape is not limited to a rectangle, and it is sufficient that the image data is continuous in both the horizontal and vertical directions. The image data of the area including the moving body 30 and the image data of the area including the stationary objects 31 and 32 are output from the partial image generation unit 114 to the image output unit 116. The image output unit 116 outputs based on the control signal of the control unit 115. When the moving body is detected, the control unit 115 outputs the area including the moving body 30 and the area including the stationary objects 31 and 32. When the moving body is not detected, the control unit 115 outputs the area including the moving body and the stationary object. Do not output the area containing.

  An example of the output format of the clipped image is shown. In this example, a case where a moving body and a stationary object of the same frame are stored as one image file will be described, but they may be stored as separate image files. The image data is saved as a RAW image in a dedicated format in a binary data format. The image data has an image data area and a header. The header includes a frame number, the number of objects, an object identification number, and coordinate information of the cut out partial image area. The frame number may be any information that can identify when the image is captured, such as the frame count from the start of imaging. Time information may be included. The number of objects indicates the number of moving objects and stationary objects that are markers. As the region information of the cut out partial image, in the case of a rectangular region, the coordinate value of the upper left corner and the coordinate value of the lower right corner are designated. The image data area stores the first partial image data as binary data and then stores the second partial image data. Store the number of moving and stationary objects. In order to determine the boundary between the first and second partial images, the number of pixels in the partial image may be calculated from the area information of the partial image. In step S <b> 24, the image storage device 12 stores the image output by the image output unit 116.

  When the above processing is performed for each frame, the image information of the area including the moving body 30 and the stationary objects 31 and 32 are detected until the moving body 30 disappears after the moving body 30 is detected as shown in FIG. The image information of the included area can be saved while maintaining the spatial resolution.

  In addition, it is not necessary to save the area including the stationary object in every frame in which the moving object is detected. An area including a stationary object may be stored only in a frame in which the moving object is first detected. This will be described with reference to FIGS. FIG. 5 shows an image processing procedure for a certain frame. In step S <b> 50, the image input unit 110 stores the image input from the difference image device 10 in the temporary storage unit 111. In step S51, the moving body detection unit 112 detects a moving body. When the moving body is detected, in step S52, the partial image generation unit 114 inputs an image from the temporary storage unit 111, cuts out an image around the moving body, and outputs the image to the image output unit 116. In step S53, the control unit 115 determines whether or not a moving object has been detected in the previous frame. The determination method can be achieved by the control unit 115 having a counter and increasing the counter when the moving body detecting unit 112 detects the moving body, and returning the counter to 0 when the moving body is not detected. . When the counter is 0, it is determined that no moving object is detected in the previous frame, and the process proceeds to step S54. In cases other than that described here, process flow proceeds to Step S55. In step S <b> 54, the stationary object detection unit 113 detects a stationary object, and the partial image generation unit 114 cuts out a region around the stationary object and outputs it to the image output unit 116. In step S <b> 55, the image storage device 12 stores the image output by the image output unit 116.

  As described above, the stationary object detection unit 113 detects the two or more stationary objects 31 and 32 only in one frame when the moving body detection unit 112 detects the moving body. With the above flow, only the first frame in which the moving object is detected is cut out and saved in the moving object peripheral area image and the stationary object peripheral area image, and in the other frames, only the moving object peripheral area image is cut out and saved. be able to. Further, the frame for storing the image of the peripheral area including the stationary object is not limited to the first frame in which the moving object is detected. Any frame from when a certain moving body is detected to disappearing may be used. The stationary object detection unit 113 may detect a stationary object in at least one frame including the moving body from the frame where the moving body is detected by the moving body detection unit 112 to the frame where the moving body disappears. If the change in brightness of a stationary object is not a concern in the frame from the detection of the moving object to the disappearance of the moving object, the above method is effective for further compressing the image data.

(Second Embodiment)
In the second embodiment of the present invention, in addition to storing image data of a peripheral region including a moving object and a peripheral region including a stationary object, position information and luminance information are added to the image data. The method will be described. The configuration of the moving body observation apparatus is the same as that of the first embodiment, but functions of position information calculation and luminance information calculation are added to the moving body detection unit 112 and the stationary object detection unit 113. The process from the detection of the moving body to the extraction of the surrounding area including the moving body and the surrounding area including the stationary object is the same as in the first embodiment. Image position information and luminance information will be described with reference to FIG.

  The position information is information indicating where the moving body 30 and / or the stationary objects 31 and 32 are located in the entire image. As the position information, the luminance barycentric coordinates of the moving body 30 and the stationary objects 31 and 32 can be used. The position information illustrated in FIG. 4 is the luminance center-of-gravity coordinates when the left end coordinate is the origin (0, 0) in an image having 1920 pixels in the horizontal direction and 1080 pixels in the vertical direction. The position information may be expressed in units of the equator and declination of the celestial coordinate system. From the information indicating the direction of the imaging device 10, the equator and declination coordinates at the left end of the image are calculated. From the angle of view of the image sensor, it can be seen how many pixel units of the image correspond to the coordinate unit on the celestial sphere. From these two pieces of information, the image coordinates and the celestial coordinate system can be converted. The luminance information is information indicating the brightness of the moving body 30 and the stationary objects 31 and 32. For example, a value obtained by averaging the count values of the digital data after A / D conversion over the area of the moving body 30 or the stationary objects 31 and 32 can be used as the luminance information. In FIG. 4, the average value of the area extracted as an object for the image having a bit depth of 14 bits is used as luminance information of the moving body 30 and the stationary objects 31 and 32. The luminance information is not limited to the average value, and may be a median value or a maximum value of an area extracted as an object. Further, as luminance information of the moving object and the stationary object, information represented by a grade may be added to the image.

  The position information and luminance information of the moving object are calculated simultaneously with the detection by the moving object detection unit 112. In addition, the position information and luminance information of the stationary object are calculated simultaneously with the detection by the stationary object detection unit 113. The information calculated by the moving object detection unit 112 and the stationary object detection unit 113 is output to the partial image generation unit 114, and then output to the image output unit 115 together with the clipped region image, and is stored by the image storage device 12. Is done. The partial image generation unit 114 adds the position information and luminance information of the moving object to the image including the moving object, and adds the position information and luminance information of the stationary object to the image including the stationary object and outputs the image.

  As an example of a method of adding position information and luminance information data, a dedicated RAW format header can be used. An area for adding position information and luminance information is secured in a predetermined header format, and the position information and luminance information are stored as binary data or text data in that portion. By adding position information and luminance information in addition to image data, it is possible to easily grasp the relative position and relative luminance of the moving body and the stationary object before analyzing the image of the image storage device 12 in detail. . As a result, the burden of observer image analysis can be reduced.

(Third embodiment)
In the first and second embodiments, the area around the moving body and the area around the stationary object are cut out from the frame where the moving body is detected to the frame where the moving body disappears. On the other hand, in the third embodiment of the present invention, an image is cut out by adding an image of a frame before detection of the moving object and a frame after the disappearance of the moving object. The reason will be described. In a moving object such as a meteor, the brightness gradually increases and the brightness gradually decreases. Therefore, there is a possibility that a weak light emission phenomenon in the middle of the movement of the moving body is captured even in the image before the moving body is detected. On the other hand, even in the image after the moving body is detected, there is a possibility that a weak light emission phenomenon that cannot be detected by the moving body detecting unit 112 due to the threshold value remains as an image. Therefore, it is necessary to cut out a portion that is assumed to have a moving body in the image before the moving body detection and the image of the moving body disappearing frame.

  The configuration of the third embodiment of the present invention is the same as that of the first and second embodiments. The image processing method will be described with reference to FIGS. FIG. 7 illustrates an image group before and after detection of a moving object. The N-1th frame is a frame image before moving object detection, the N frame is a moving object detection image, and the N + 3th frame is a moving object lost image. Since the image processing while the moving body 30 is detected is the same as the method of the first to third embodiments, the image extraction processing of the pre-moving body detection frame and the moving body disappearing frame will be described.

  FIG. 8A shows a flowchart for extracting the moving object estimated position in the frame immediately before the moving object detection. First, in step S80, the moving object detection unit 112 extracts the position coordinates of the moving object 30 in the moving object detection frame (Nth frame) and the next frame (N + 1th frame). The position coordinates of the moving body 30 may be, for example, the luminance center-of-gravity coordinates of the moving body 30.

  In step S81, the moving object detection unit 112 calculates the difference between the position coordinates of the moving object 30 between the moving object detection frame (Nth frame) and the next frame (N + 1th frame), and the moving object detection frame (N The estimated position coordinates of the moving body 33 in the (-1 frame) image are calculated. For example, the luminance centroid coordinates of the moving body 30 in the Nth frame are (X, Y) = (254, 167), and the luminance centroid coordinates of the moving body 30 in the N + 1th frame are (X, Y) = (356, 268). And In this case, the difference coordinates are (ΔX, ΔY) = (102, 101), and the estimated position coordinates in the (N−1) th frame are (254-102, 167-101) = (152, 66).

  In step S82, the moving body detection unit 112 determines whether the moving body estimated position of the moving body 33 is within the image area. In the above example, since the coordinates are in the image area, the process proceeds to step S83. If it is out of the image area, the process is terminated at that time.

  In step S <b> 83, the partial image generation unit 114 cuts out a peripheral region including the estimated moving body position of the moving body 33 in the image of the frame immediately before the moving body detection (N−1 frame). In step S84, the image storage device 12 stores the image cut out by the partial image generation unit 114.

  Through the above image processing procedure, an image of the area around the estimated position of the pre-moving object detection frame is acquired. Also in the moving object disappearance frame (N + 3th frame), the estimated moving object position of the moving object 34 is calculated from the frame before the moving object disappearance (N + 2 frame) and the image before the moving object disappearance 2 frames (N + 1 frame). . If the estimated position of the moving object is within the image area, an area around the estimated position in the moving object disappearing frame is cut out.

  FIG. 8B shows a flowchart when the moving object disappears. In step S85, the moving body detection unit 112 acquires the luminance gravity center position information of the moving body 30 at the position immediately before the disappearance (N + 2 frame in FIG. 7) and 2 frames before the disappearance (N + 1 frame). In step S86, the moving object detection unit 112 calculates the estimated moving object position of the moving object 34 in the moving object disappearing frame (N + 3rd frame) from the difference between the positions of the two frames. In step S87, the moving body detection unit 112 determines whether or not the moving body estimated position is within the image area. If the estimated moving body position is within the image area, in step S88, the partial image generating unit 114 cuts out an area including the estimated moving body position of the moving object 34 in the moving object disappearing frame (N + 3th frame). In step S89, the image storage device 12 stores the image cut out by the partial image generation unit 114.

  As described above, the partial image generation unit 114 determines the estimated positions of the moving objects 33 and 34 in the frame before the frame where the moving object is detected by the moving object detection unit 112 and in the frame where the moving object disappears. The area of the estimated position of the moving object is cut out and output. By performing the above processing, it is possible to cut out a moving body estimated image before the moving body is detected and when the moving body is lost. Further, as in the second embodiment, the position information and luminance information of the moving body may be stored. With respect to the frame before moving object detection and the frame after moving object detection, the position information of the estimated position of the moving object and, at the same time, the image of the peripheral area of the stationary object are stored, so that the moving object detection unit 112 cannot detect it. The phenomenon can be analyzed later with detailed image analysis.

(Fourth embodiment)
In the third embodiment, the image of the frame before the moving object detection and the image after the moving object disappearance frame are additionally cut out. In the flowchart of the third embodiment, when two or more moving objects are detected, it is not possible to determine whether the moving objects are the same object, and there is a possibility that the estimated moving object position is incorrect. Therefore, in the fourth embodiment of the present invention, the moving object estimated position of the frame before the moving object detection is calculated by using the moving object estimated position as the moving object detection frame and the subsequent two frames. The calculation method will be described.

  FIG. 9 shows a group of images before and after detecting the moving object. A moving body 91 appears with a delay of one frame of the moving body 90. In this case, when the N + 1 frame and the N frame are compared, the mobile body estimated position can be correctly calculated by comparing the mobile bodies 90 with each other. However, if the mobile body 91 of N + 1 frame and the mobile body 90 of N frame are compared, the mobile body estimated position of the mobile body 92 in the N-1 frame cannot be calculated correctly.

  Therefore, in the fourth embodiment, it is determined whether or not the moving body is the same from the position coordinates and luminance values of the N and N + 1 frames, and the moving body estimated position of the moving body 92 is calculated. This will be described with reference to the schematic diagram of the moving object continuous image of FIG. FIG. 10 is a diagram showing the position coordinates and luminance count values of the moving body 90 and the moving body 91. The brightness count values of N + 1 frame and N frame are compared. For the count value 50 of N frame, 48 close to the brightness count value in N + 1 frame is regarded as the same object, and the moving body estimated position of the moving body 92 in N-1 frame is calculated from the difference coordinate value. Then, the partial image generation unit 114 cuts out a region around the moving body estimated position of the moving body 92. In the partial image generation unit 114, the moving body detection unit 112 detects the moving body based on the frame where the moving body is detected by the moving body detection unit 112 and the detection position and luminance information of the moving body of at least one frame thereafter. The moving object estimated position of the moving object in the previous frame is calculated.

(Fifth embodiment)
In the fifth embodiment of the present invention, in the mobile object estimated position calculation of the fourth embodiment, the mobile object moves at substantially the same speed using three frames of N frames, N + 1 frames, and N + 2 frames. Assuming that the moving object estimated position is calculated. The method of the fifth embodiment can be used when the brightness count values of the two moving objects are close. FIG. 11 shows the coordinates of the moving body 90 and the moving body 91 on the image. The moving body detection unit 112 calculates difference coordinates Δv1 and Δv2 between the moving body of the N + 1 frame and the N frame. Δv1 = (200, 50), Δv2 = (100, 400). Next, difference coordinates Δv11, Δv22, Δv21, Δv12 of the moving body of the N + 2 frame and the N + 1 frame are calculated. Δv11 = (200, 50), Δv22 = (300, 0), Δv21 = (300, −300), Δv12 = (200, 350). Since the difference coordinates of Δv1 and Δv11 are the same, the speed of the moving body 90 that appears in the N frame is Δv1 = (200, 50), and the estimated moving position of the moving body 92 in the N−1 frame is (250−200, 100). −50) = (50, 50). Then, the partial image generation unit 114 cuts out the peripheral area of the coordinates.

  The partial image generation unit 114 determines the frame of the frame in which the moving body is detected by the moving body detection unit 112 based on the detection position of the moving body in which the moving body is detected by the moving body detection unit 112 and at least two frames thereafter. The moving body estimated position of the moving body in the previous frame is calculated.

  According to the first to fifth embodiments, in the moving object observation, while compressing the capacity of the image data stored in the image storage device, without reducing the information amount of the area of the moving object and the desired stationary object, Image data can be stored efficiently.

  The above-described embodiments are merely examples of implementation in carrying out the present invention, and the technical scope of the present invention should not be construed in a limited manner. That is, the present invention can be implemented in various forms without departing from the technical idea or the main features thereof.

DESCRIPTION OF SYMBOLS 10 Imaging device, 112 Moving body detection part, 113 Still object detection part, 114 Partial image generation part

Claims (6)

An imaging device for imaging a plurality of consecutive frames;
A moving body detection unit that detects a moving body based on the plurality of captured frames;
A stationary object detection unit that detects a stationary object based on the captured frame;
A partial image generation unit that cuts out the areas of the detected moving object and stationary object,
The stationary object detection unit detects a stationary object in at least one frame including the moving body from a frame where the moving body is detected by the moving body detection unit to a frame where the moving body disappears. Mobile observation device.   The partial image generation unit adds position information and luminance information of the moving object to an image including the moving object, and adds position information and luminance information of the stationary object to an image including the stationary object and outputs the image. The moving body observation apparatus according to claim 1.   The partial image generation unit calculates a moving body estimated position of the moving body in a frame before the frame in which the moving body is detected by the moving body detection unit and in a frame in which the moving body has disappeared. 3. The moving object observation apparatus according to claim 1, wherein a region of the estimated position is cut out and output.   The partial image generation unit detects the moving object by the moving object detection unit based on the frame where the moving object is detected by the moving object detection unit and the detection position and luminance information of the moving object of at least one frame thereafter. 4. The moving object observation apparatus according to claim 3, wherein a moving object estimated position of the moving object in a frame before the frame is calculated.   The partial image generation unit is configured to detect a frame in which the moving object is detected by the moving object detection unit based on a detection position of the moving object detected by the moving object detection unit and at least two frames after that. The moving body observation apparatus according to claim 3, wherein a moving body estimated position of the moving body in a previous frame is calculated.   The said stationary object detection part detects two or more stationary objects only in 1 frame when a mobile body is detected by the said mobile body detection part, The any one of Claims 1-5 characterized by the above-mentioned. The moving body observation apparatus described.

Images