Information processing system and information processing method

Abstract

To perform monitoring in a field that requires high security without significantly increasing cost, using captured image data.SOLUTION: An information processing unit 51-1 of a fail-safe computer 33, serving as an information processing unit, includes an output processing unit 96 which is configured to: compare, in two or more systems, a first CRC added to an image by a data processing unit 23, a second CRC calculated by a first CRC calculation processing unit 72-1, and a third CRC calculated by a second CRC calculation processing unit 72-2; and detect and output a failure of one of an imaging processing unit 11, a first image processing unit 41, and a second image processing unit 42.SELECTED DRAWING: Figure 7

Description

[Technical field]

[0001] The present invention relates to an information processing system and an information processing method. [Background technology]

[0002] Various monitoring systems are widely used to ensure safety. Most of the general monitoring systems are backups for human monitoring, and are acceptable if they have a certain degree of reliability, or are used for monitoring to improve maintainability and availability, and even if an abnormality is overlooked, it does not directly progress to a dangerous event. In contrast, when there is a possibility that an abnormality in the monitored object may progress to a dangerous event (an event that may cause harm to people or property) if an abnormality in the monitored object is overlooked due to a system failure, etc., even higher safety is required. For example, a system that detects whether a person remains on a railroad crossing when a train approaches, or a system that monitors the presence or absence of obstacles in front of a traveling train and reflects the results in the control of the train, etc., may progress to a dangerous event if an abnormality in the monitored object is overlooked due to a system failure, etc.

[0003] In other words, monitoring systems in the railway signaling field must ensure high safety. In this field, the methods for ensuring safety that have been conventionally used can be roughly divided into two categories.

[0004] One of them is to configure the system using elements with asymmetric error characteristics. For example, a signal relay has a significantly different probability of failure on the operating side (ON side) and the recovery side (OFF side), and can be considered to transition to the recovery side in the event of a failure. Therefore, in railway signaling systems, the safe side of the control state (input / output) is defined, and the circuit is designed by allocating the safe side to the recovery side of the signal relay, ensuring safety in the event of a failure.

[0005] The other is to constantly monitor the system for faults and immediately fix to the safe side when a fault is detected. Electronic devices and electronic elements, such as computers, do not have the asymmetric error characteristics described above, and it is not possible to identify the state when a fault occurs. Therefore, safety can be ensured by a method of constantly monitoring for faults and immediately fixing to the safe side when a fault is detected.

[0006] In order to ensure safety using this method, it is necessary that the outputs of all components such as the system and subsystems are all correct, i.e., safe, until a fault is detected, that an output on the safe side is provided when a fault is detected, and that after a fault is detected and an output on the safe side is provided, that state is fixed reliably without the need to resume control with some kind of trigger.

[0007] To meet this requirement, a system configuration is used in which the same process is run in two or more systems, and the process results of the two or more systems are compared to detect errors. In this method, the two or more systems are configured independently, and it is required that common cause failures be negligible and that comparison be performed at a frequency high enough to account for the failure probability. In other words, if a common cause failure occurs in two or more systems, an error cannot be detected even by comparison. Also, if a failure occurs in one system after a failure occurs in the other system before the state is fixed to the safe side, there may be cases where an error cannot be detected even by comparing the process results of each system.

[0008] Thus, in monitoring systems for railway signals, it is necessary to check that the detection function is sound by running the same process in two or more systems, i.e., by redundancy. For this reason, various information processes are carried out by information processing devices called fail-safe computers, which execute the same process in two or more systems in terms of both software and hardware, and determine that the processing results are normal only when the processing results are identical.

[0009] Furthermore, with recent advances in image processing and image recognition technology, it is expected that systems using camera images will be applied to surveillance.

[0010] For example, in a vehicle display system, in order to detect freezing of images captured by two cameras, there is a technique for comparing images to be compared using a sub-microcomputer and an additional microcomputer different from the main microcomputer that performs display processing. When it is detected that there is a change in the image captured by one camera but there is no change in the image captured by the other camera, a process such as resetting the main microcomputer is executed. Also, here, there is a technique for detecting whether there is a change in the images by a method such as comparing hash values ​​(e.g., SHA, MD5, CRC) in small areas between the images to be compared (e.g., Patent Document 1). [Prior art documents] [Patent documents]

[0011] [Patent Document 1] JP 2018-79839 A Summary of the Invention [Problem to be solved by the invention]

[0012] However, the monitoring system using the technology described in Patent Document 1 is a monitoring system for improving maintainability and availability, and even if an error is overlooked, it does not directly progress to a dangerous event. Normally, it is impossible to uniquely assign safe side or dangerous side to the data of a camera image. In addition, almost all of the components inside a camera, such as the image sensor, do not have asymmetric error characteristics. For this reason, when attempting to monitor a field requiring high safety using the technology described in Patent Document 1, it is necessary to configure a system in which the same processing is operated in two or more systems, and information from the two or more systems is collated to detect errors.

[0013] When using the technology described in Patent Document 1 to monitor fields that require high safety, it is necessary to verify that the "detection function is sound" through redundancy, so the imaging system and processing system must be made redundant as a set. In that case, two or more imaging systems are required, which is expensive in terms of both parts cost and installation cost. Even if a system is constructed using two or more imaging systems, the images obtained by the multiple imaging systems are different from each other, so they cannot be directly compared, making it difficult to detect failures in the processing system.

[0014] Furthermore, although the above-mentioned fail-safe computer can ensure safety, because it has a redundant configuration, in order to process images obtained from two or more imaging systems at a sufficient speed, it is necessary to build a very high-performance computer, which significantly increases costs.

[0015] Therefore, an object of the present invention is to solve the above-mentioned problems, that is, to provide an information processing system and an information processing method that can utilize captured image data to monitor fields where high security is required without significantly increasing costs. [Means for solving the problem]

[0016] One aspect of an information processing system of the present invention is an information processing system including an imaging unit, a first image processing unit that processes an image captured by the imaging unit, a second image processing unit that executes the same processing as the first image processing unit, and an information processing unit that executes processing for managing the operation of the imaging unit, the first image processing unit, and the second image processing unit in two or more systems, in which the imaging unit has a detection information attachment processing unit that calculates a first CRC based on the captured image and attaches it to frame data of the image, and the first image processing unit has a first image acquisition unit that acquires an image to which the first CRC has been attached by the detection information attachment processing unit, a first CRC calculation processing unit that calculates a second CRC for the image acquired by the first image acquisition unit, a first image processing execution unit that executes processing on the image acquired by the first image acquisition unit, and a first CRC attached to the image acquired by the first image acquisition unit, the second CRC calculated by the first CRC calculation processing unit, and a processing result by the first image processing execution unit. the second image processing unit has a second image acquisition unit that acquires an image to which a first CRC has been added by the detection information addition processing unit, a second CRC calculation processing unit that calculates a third CRC for the image acquired by the second image acquisition unit, a second image processing execution unit that executes processing on the image acquired by the second image acquisition unit, and a second transmission unit that sends the first CRC added to the image acquired by the second image acquisition unit, the third CRC calculated by the second CRC calculation processing unit, and the processing result by the second image processing execution unit to the information processing unit, and the information processing unit has a fault detection unit that compares the first CRC added to the image by the detection information addition processing unit, the second CRC calculated by the first CRC calculation processing unit, and the third CRC calculated by the second CRC calculation processing unit in two or more systems, and detects a fault in any of the imaging unit, the first image processing unit, or the second image processing unit based on the comparison result.

[0017] Another aspect of the information processing system of the present invention is characterized in that the imaging unit further has a test pattern supply control unit that controls the supply of the test pattern, the detection information addition processing unit adds a first CRC to the test pattern whose supply is controlled by the test pattern supply control unit, the first CRC calculation processing unit of the first image processing unit further calculates a second CRC for the test pattern, and the second CRC calculation processing unit of the second image processing unit further calculates a third CRC for the test pattern, the information processing unit further has a CRC memory unit that stores a correct value of the CRC calculation result for the test pattern whose supply is controlled by the test pattern supply control unit, and the fault detection unit compares the first CRC added to the test pattern by the detection information addition processing unit, the second CRC corresponding to the test pattern calculated by the first CRC calculation processing unit, and the third CRC corresponding to the test pattern calculated by the second CRC calculation processing unit with the correct values ​​stored in the CRC memory unit, and detects a fault in any of the imaging unit, the first image processing unit, or the second image processing unit based on the comparison result.

[0018] Another aspect of the information processing system of the present invention is characterized in that the first image processing unit has a first memory unit that stores a first background model for detecting an object by background subtraction, and a first background subtraction calculation unit that performs background subtraction on the image and test pattern supplied from the imaging unit using the first background model stored in the first memory unit, and the second image processing unit has a second memory unit that stores a second background model that is identical to the first background model and for detecting an object by background subtraction, and a second background subtraction calculation unit that performs background subtraction on the image and test pattern supplied from the imaging unit using the second background model stored in the second memory unit.

[0019] In another aspect of the information processing system of the present invention, the detection information adding processing unit of the imaging unit adds a first CRC to a predetermined position in frame data of a captured image.

[0020] In another aspect of the information processing system of the present invention, the detection information adding processing section further adds a clock counter value to a predetermined position in frame data of the captured image.

[0021] In another aspect of the information processing system of the present invention, the detection information adding processing section further adds a frame counter value to a predetermined position within frame data of the captured image.

[0022] Another aspect of the information processing system of the present invention is characterized in that the first sending unit of the first image processing unit further sends out the frame counter value assigned by the detection information adding processing unit, and the second sending unit of the second image processing unit further sends out the frame counter value assigned by the detection information adding processing unit, and the information processing unit further has an image processing buffer that holds multiple frames of the processing result by the first image processing execution unit and the frame counter value sent out by the first sending unit, as well as the processing result by the second image processing execution unit and the frame counter value sent out by the second sending unit, and a comparison unit that refers to the processing result by the first image processing execution unit held in the image processing buffer and the frame counter value assigned to the processing result by the second image processing execution unit, and compares the processing result by the first image processing execution unit and the processing result by the second image processing execution unit corresponding to images of the same frame.

[0023] One aspect of an information processing method of the present invention is an information processing method for an information processing system including an imaging unit, a first image processing unit that processes an image captured by the imaging unit, a second image processing unit that executes the same processing as the first image processing unit, and an information processing unit that executes processing for managing operations of the imaging unit, the first image processing unit, and the second image processing unit in two or more systems, the information processing method including a first CRC calculation step of calculating a first CRC based on an image captured by the imaging unit, a detection information adding step of adding the first CRC calculated by the processing of the first CRC calculation step to frame data of the image, a first image acquisition step of acquiring an image to which the first CRC has been added, a second image acquisition step of acquiring an image to which the first CRC has been added, a second CRC calculation step of calculating a second CRC for the image acquired by the processing of the first image acquisition step, and a third CRC calculation step of calculating a third CRC for the image acquired by the processing of the second image acquisition step in the second image processing unit. a first image processing step of processing an image acquired by the processing of the first image acquisition step in a first image processing unit; a second image processing step of processing an image acquired by the processing of the second image acquisition step in a second image processing unit; a first sending step of sending, to an information processing unit, the first CRC added to the image, the second CRC calculated by the processing of the first CRC calculation step, and a processing result by the processing of the first image processing step in the first image processing unit; The image processing method includes a second transmission step of transmitting the first CRC provided, the third CRC calculated by the processing of the second CRC calculation step, and the processing result of the processing of the second image processing step to an information processing unit, and a fault detection step of comparing the first CRC, the second CRC, and the third CRC transmitted by the processing of the first transmission step and the second transmission step in two or more systems in the information processing unit, and detecting a fault in any of the imaging unit, the first image processing unit, or the second image processing unit based on the comparison result. Effect of the Invention

[0024] According to the present invention, captured image data can be used to monitor fields where high security is required without significantly increasing costs. [Brief description of the drawings]

[0025] [Figure 1] FIG. 2 is a functional block diagram showing a functional configuration of the image processing system 1. [Diagram 2] 2 is a functional block diagram showing a functional configuration of a data processing unit 23. FIG. [Diagram 3] 13 is a diagram for explaining an image data frame to which a frame counter value, a clock counter value, and a CRC calculated value have been added. FIG. [Figure 4] 2 is a functional block diagram showing a functional configuration of a first image processing unit 41 of an image processing unit 32. FIG. [Diagram 5] FIG. 13 is a diagram showing an example of the functional configuration of a first object detection processing unit 73-1 when a background subtraction method is used. [Figure 6] FIG. 13 is a diagram for explaining the processing of a first image data reduction processing unit 75-1. [Figure 7] FIG. 2 is a functional block diagram for explaining functions of an information processing unit 51-n. [Figure 8] 1 is a diagram for explaining an image comparison process using the image processing buffer 101. FIG. [Figure 9] 1 is a diagram for explaining an image comparison process using the image processing buffer 101. FIG. [Figure 10] 11 is a diagram for explaining the processing of the Hamming distance test processing unit 103. FIG. [Figure 11] 11 is a diagram for explaining the processing of the Hamming distance test processing unit 103. FIG. [Figure 12] 4 is a flowchart for explaining processing of the imaging processing unit 11. [Figure 13] 10 is a flowchart for explaining the processing of a first image processing unit 41 and a second image processing unit 42. [Figure 14] 13 is a flowchart for explaining a CRC check process of an information processing unit 51-n. [Figure 15] 13 is a flowchart for explaining image processing result comparison processing of the information processing unit 51-n. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0026] Hereinafter, an information processing apparatus, an information processing method, an information processing system, and a program according to an embodiment of the present invention will be described with reference to the drawings.

[0027] FIG. 1 is a functional block diagram showing the functional configuration of an image processing system 1 according to an embodiment of the present invention.

[0028] The image processing system 1 includes an imaging processing unit 11 and a judgment processing unit 12. The image processing system 1 may be configured by one device or may be configured by multiple devices. The imaging processing unit 11 is installed in a position where it can capture images used for railway safety, such as inside a railroad crossing or around a platform. Here, the imaging processing unit 11 is configured as one device, and the judgment processing unit 12 is configured as one device.

[0029] Here, the imaging processing unit 11 is described as being provided at a position where it can acquire images for monitoring the inside of the railroad crossing, and performs imaging processing of predetermined locations, as well as executing predetermined processing for determining whether or not any malfunction has occurred in the image processing system 1. The judgment processing unit 12 is described as being provided at a position where it can acquire images for monitoring the inside of the railroad crossing, and performs predetermined processing for determining whether or not any malfunction has occurred in the image processing system 1, as well as executing image processing required for safety monitoring processing, specifically, for example, processing for determining whether or not an object such as a person or a car exists inside the railroad crossing, based on the images captured by the imaging processing unit 11, as well as determining whether or not any malfunction has occurred in the image processing system 1.

[0030] The imaging processing unit 11 is configured to include the functions of a control unit 21, an imaging unit 22, a data processing unit 23, and a transmission side interface processing unit 24. The imaging processing unit 11 has only one imaging unit 22, one data processing unit 23, and one transmission side interface processing unit 24. If two of each of these are provided to provide a redundant circuit configuration, a failure can be found by comparing the respective processing results. However, this not only increases costs, but also causes problems such as the images of the target area captured by different imaging units being different and therefore not being directly collated, and the processing timing of the image processing being not easily synchronized completely, making it difficult to detect failures in the processing system. Therefore, the image processing system 1 has a function that allows a failure that can be a common cause failure to be reliably detected at a later stage of processing when the failure occurs in the imaging processing unit 11.

[0031] The control unit 21 controls each part of the imaging processing unit 11 .

[0032] The imaging unit 22 includes an imaging element and an A / D converter, and performs imaging processing at a predetermined frame rate and supplies the results to the data processing unit .

[0033] The data processing unit 23 executes conventional processing for performing necessary processing (such as processing for correcting variations between pixels) on the output image of the imaging unit 22. Furthermore, in addition to these conventional processing, the data processing unit 23 executes processing required for accurately detecting failures in each part of the image processing system 1.

[0034] That is, since the imaging processing unit 11 shown in FIG. 1 has only one imaging section 22, one data processing section 23, and one transmission side interface processing section 24, each of these sections may have a common cause failure. For this reason, it is necessary that failures of each section of the imaging processing unit 11 can be detected in a subsequent stage. Also, in order to detect failures of each section in a subsequent stage, it is necessary to insert necessary information in a previous stage of processing of captured image data. For this reason, the data processing section 23 performs processing required to accurately detect failures of each section of the image processing system 1, in addition to processing such as correction of captured images similar to conventional processing. The detailed functional configuration of the data processing section 23 will be described later with reference to FIG. 2.

[0035] The transmission side interface processing unit 24 adds a frame check sequence (hereinafter referred to as FCS) for checking for transmission errors to the information including the captured image data supplied from the data processing unit 23, and supplies the information to the determination processing unit 12.

[0036] Next, the judgment processing unit 12 executes a predetermined process from the captured image, detects information necessary for the safety of railway signals, judges the state, and judges whether the image processing system 1 is operating correctly. The judgment processing unit 12 includes a receiving side interface processing unit 31, an image processing unit 32, and a fail-safe computer 33.

[0037] The fail-safe computer 33 has a redundant configuration, and therefore can ensure safety, but its processing speed is relatively slow. That is, it is difficult for the fail-safe computer 33 to process images obtained from two or more imaging systems at a sufficient speed, or to perform all of the processes required to accurately detect failures in each part of the image processing system 1 (hereinafter referred to as failure detection processes) at a sufficiently high frequency against the failure probability. For this reason, the judgment processing unit 12 is provided with an image processing unit 32 in addition to the fail-safe computer 33, and the image processing and part of the failure detection process are executed by the image processing unit 32. The image processing unit 32 and the fail-safe computer 33 operate asynchronously even if they operate at the same cycle.

[0038] Next, the processing of each part of the decision processing unit 12 will be described.

[0039] The receiving interface processor 31 acquires information including the captured image data supplied from the imaging processing unit 11, executes an FCS test, and if there is no problem, supplies the captured image data supplied from the imaging processing unit 11 to each of the first image processor 41 and the second image processor 42 of the image processing unit 32. Furthermore, for example, when an image captured on a monitor or the like is displayed on a monitor (not shown) or stored in a memory (not shown), the receiving interface processor 31 outputs the captured image data supplied from the imaging processing unit 11 to a corresponding device or the like, but a description of these processes will be omitted here.

[0040] The image processing unit 32 has a first image processing section 41 and a second image processing section 42 having the same functions, that is, has a redundant configuration with two identical systems. The first image processing section 41 and the second image processing section 42 of the image processing unit 32 execute image processing on captured image data. Furthermore, the first image processing section 41 and the second image processing section 42 execute a part of the fault detection process based on various information added by the data processing section 23. The detailed functions of the first image processing section 41 and the second image processing section 42 will be described later with reference to Figs. 4 and 5.

[0041] The fail-safe computer 33 is an information processing device that operates one process in two or more systems in terms of software and hardware, and determines that a normal processing result has been obtained only when the respective processing results are identical. The fail-safe computer 33 is configured to include the functions of n information processing units 51-n (n≧2), a processing result determination unit 52, and a processing result output control unit 53.

[0042] The information processing unit 51-n judges whether or not the image processing results by the first image processing unit 41 and the second image processing unit 42 of the image processing unit 32 are the same for each processing system, and supplies the judgment result to the processing result judgment unit 52. Furthermore, the information processing unit 51-n executes fault detection processing in each processing system based on a part of the processing results of the fault detection processing by the first image processing unit 41 and the second image processing unit 42 of the image processing unit 32, and supplies the result to the processing result judgment unit 52. Details of the functions of the information processing unit 51-n will be described later with reference to FIG. 7.

[0043] If the image processing results are the same in the information processing units 51-n and no fault is detected by the fault detection process, the processing result determination unit 52 determines that the image processing results by the first image processing unit 41 and the second image processing unit 42 of the image processing unit 32 are correct, generates a control signal for executing various processes based on the result, and supplies it to the processing result output control unit 53. If the image processing results are not the same in all processing systems or if a fault is detected by the fault detection process, the processing result determination unit 52 controls various operations to the safety side; specifically, for example, it generates a control signal for instructing a signal operation to prevent a train from entering the corresponding railroad crossing, and supplies it to the processing result output control unit 53.

[0044] The processing result output control unit 53 controls the output of the control signal supplied from the processing result determination unit 52 to a predetermined other device, such as a management control device that manages railway signals.

[0045] 2 is a functional block diagram showing a more detailed functional configuration of the data processing unit 23. As described above, the data processing unit 23 performs processes required for accurately detecting failures in each unit of the image processing system 1, in addition to the conventional processes such as correction of captured images. The data processing unit 23 is configured to include the functions of an image data acquisition unit 61, a test pattern generation unit 62, a correction processing unit 63, a CRC generation unit 64, a frame counter 65, a clock counter 66, and an image data frame generation unit 67.

[0046] In the data processing unit 23, the imaging data acquisition unit 61 and the correction processing unit 63 have the same functions as those of the data processing unit 23 of a conventional general camera device. When the conventional data processing unit 23 is configured with an FPGA (Field-Programmable Gate Array), it is possible to easily add the functions of the test pattern generation unit 62, the CRC generation unit 64, the frame counter 65, the clock counter 66, and the image data frame generation unit 67 to the conventional data processing unit 23.

[0047] The imaging data acquisition unit 61 acquires the frame image data supplied from the imaging unit 22 , and supplies it to a correction processing unit 63 .

[0048] The test pattern generating unit 62 generates a test pattern at a predetermined timing and supplies it to the correction processing unit 63. When a background subtraction method is used in the object detection processing unit described later, it is preferable that the test pattern generated by the test pattern generating unit 62 has a sufficient difference in pixel value from the background image. The test pattern is generated at an interval that is sufficiently high compared to the failure probability. The test pattern may be generated when a request is received from the judgment processing unit 12.

[0049] The correction etc. processing unit 63 executes conventional image correction processing etc., such as processing to correct variations between pixels of the frame image data supplied from the imaging data acquisition unit 61, and supplies the image data that has been subjected to correction etc. to the image data frame generation unit 67 and the CRC generation unit 64. The correction etc. processing unit 63 also inserts the supplied test pattern into the frame image data and supplies it to the image data frame generation unit 67.

[0050] Also, a test pattern generating unit 62 may be provided in the imaging unit 22, and the generated test pattern may be inserted in a stage preceding the A / D converter of the imaging unit 22, or the test pattern may be forcibly captured using an electronic shutter or the like. The earlier the test pattern is inserted, the more abnormalities in elements and the like can be detected. However, when a conventional data processing unit is configured with an FPGA, it is easy to add a function for inserting a test pattern into frame image data in a stage preceding the correction etc. processing unit 63, so here, an explanation will be given of the insertion of a test pattern in the configuration shown in FIG. 2.

[0051] The test pattern inserted into the captured image is used to detect the occurrence of an abnormality in the function of detecting image fixation using CRC, which will be described later, and the test pattern inserted into the captured image is also used to detect the occurrence of an abnormality in object detection using the background subtraction method, which will be described later.

[0052] The CRC generation unit 64 generates a CRC (Cyclic Redundancy Check) based on the frame image data supplied from the correction etc. processing unit 63 and supplies it to the image data frame generation unit 67. The CRC generation unit 64 calculates a CRC based on information other than the pixel positions where information other than image data is embedded by the image data frame generation unit 67 described later. The CRC value calculated by the CRC generation unit 64 is referred to as a first CRC value. By executing a CRC test in the determination processing unit 12 described later, it becomes possible to detect fixation of image data due to a failure of the imaging element or A / D converter in the imaging unit 22, a failure of the reading circuit of the correction etc. processing unit 63, etc.

[0053] It is assumed here that even when capturing an image of a still object, minute fluctuations occur on a pixel-by-pixel basis, and the calculated CRC value is not always the same. In contrast, in the case of the imaging unit 22, which does not generate minute fluctuations, it is necessary to set the time element according to the frequency of movement of the object to be captured.

[0054] Furthermore, even if an abnormality occurs in the image data in the transmission path, it can be detected by the CRC test. As described above, in the communication between the imaging processing unit 11 and the judgment processing unit 12, the transmission error check using FCS is performed, but the transmission error check using FCS itself is performed only in one system, and the CRC test is effective from the viewpoint of the necessity of ensuring the consistency of the data input to the two systems of image processing, i.e., the first image processing unit 41 and the second image processing unit 42, by detecting a failure in the transmission path, etc., from the end of the transmission error check using FCS to the execution of image processing.

[0055] The frame counter 65 is a counter that counts the frame number of the frame image data in accordance with the frame rate of the imaging unit 22. The clock counter 66 is a counter that counts clock pulses. The values ​​of these counters are used to detect frame missing or frame rate abnormality during image processing. The image processing unit 32 and the fail-safe computer 33, which will be described later, operate asynchronously even if they operate at the same cycle, and the first image processing unit 41 and the second image processing unit 42 operate asynchronously. Therefore, the counter value of the frame counter 65 is used to correctly detect the agreement of the image processing performed in the first image processing unit 41 and the second image processing unit 42.

[0056] The image data frame generation unit 67 generates an image data frame by adding the frame counter value, the clock counter value, and the first CRC value to the frame image data including the test pattern supplied from the correction etc. processing unit 63, and supplies it to the transmitting side interface processing unit 24.

[0057] With reference to FIG. 3, an image data frame to which the frame counter value, clock counter value, and CRC calculated value have been added will be described.

[0058] In the communication between the imaging processing unit 11 and the judgment processing unit 12, if the data frame configuration is defined by a standard, a dedicated data frame cannot be provided, and if a data frame in which a frame counter value, a clock counter value, and a first CRC value are written is newly provided in addition to the image data frame, the image communication speed may be slowed down, or it may become necessary to distinguish the image data frame from other data frames in the subsequent image processing. Therefore, the image data frame generating unit 67 describes the frame counter value, the clock counter value, and the first CRC value using a data area of ​​several pixels of the data frame defined by the standard.

[0059] 3, the frame counter value, clock counter value, and CRC calculated value are written to the data area of ​​three pixels, (637,479), (638,479), and (639,479), of the 639×479 pixel image data area. The frame counter value, clock counter value, and first CRC value are preferably written in locations that are unlikely to interfere with the object detection process described below, such as the edges or corners of the image.

[0060] 4 is a functional block diagram showing the functional configuration of the first image processing section 41 of the image processing unit 32. The first image processing section 41 is configured to include the functions of a first image data acquisition section 71-1, a first CRC calculation processing section 72-1, a first object detection processing section 73-1, a first clock check counter 74-1, a first image data reduction processing section 75-1, and a first information transmission processing section 76-1.

[0061] The second image processing unit 42 is configured to include the functions of a second image data acquisition unit 71-2, a second CRC calculation processing unit 72-2, a second object detection processing unit 73-2, a second clock check counter 74-2, a second image data reduction processing unit 75-2, and a second information transmission processing unit 76-2, and since these perform the same processing as the first image data acquisition unit 71-1, the first CRC calculation processing unit 72-1, the first object detection processing unit 73-1, the first clock check counter 74-1, the first image data reduction processing unit 75-1, and the first information transmission processing unit 76-1, they will not be illustrated and will not be described in detail.

[0062] The fail-safe computer 33 downstream of the image processing unit 32 is an information processing device that executes processing in two or more systems in terms of software and hardware, and determines that a normal processing result has been obtained only when the respective processing results are identical, and therefore has a relatively slow processing speed. However, in order to reliably detect a failure in the monitoring process, it is required to perform matching at a frequency sufficiently high relative to the failure probability. Therefore, the first image processing section 41 and the second image processing section 42 of the image processing unit 32 perform image processing for detecting whether or not the captured image data contains objects other than the background (such as people or cars), and reduction processing of the processing results, and also perform CRC calculation processing of the acquired image as part of the processing required for CRC inspection.

[0063] The first image data acquisition unit 71-1 acquires frame image data including the frame counter value, clock counter value, and first CRC value information described with reference to Fig. 3 from the receiving side interface processing unit 31, extracts the frame counter value and the first CRC value, and supplies them to the first information transmission processing unit 76-1. The first image data acquisition unit 71-1 also supplies the frame image data including the frame counter value, clock counter value, and first CRC value information to the first CRC calculation processing unit 72-1, the first object detection processing unit 73-1, and the first clock check counter 74-1.

[0064] The first CRC calculation processing unit 72-1 calculates a CRC based on information other than the pixel positions in which information other than image data is embedded as described with reference to Fig. 3 by processing of the image data frame generation unit 67 out of the frame image data supplied from the first image data acquisition unit 71-1, and supplies the calculation result to the first information transmission processing unit 76-1. Here, the CRC calculation value generated by the first CRC calculation processing unit 72-1-1 is referred to as the 2a CRC value (the CRC calculation value generated by the second CRC calculation processing unit 72-2 is referred to as the 2b CRC value).

[0065] The first object detection processing unit 73-1 detects an abnormality within the imaging range using the frame image data supplied from the first image data acquisition unit 71-1. Specifically, for example, the first object detection processing unit 73-1 executes image processing for detecting whether or not an object such as a person or a car exists within the railroad crossing, and supplies the image processing result (i.e., the detection result of the presence or absence of an object) to the first information transmission processing unit 76-1, and supplies the processed image data to the first image data reduction processing unit 75-1. The image processing method for object detection by the first object detection processing unit 73-1 may be any method, but here, the description will be given assuming that the background difference method is used. Details of the first object detection processing unit 73-1 using the background difference method will be described later with reference to FIG. 5. When the first object detection processing unit 73-1 uses the background difference method, the processed image data becomes binarized image data.

[0066] The first clock check counter 74-1 is a counter for clock checking, which extracts a clock counter value contained in the frame image data supplied from the first image data acquisition unit 71-1, and supplies it to the first information transmission processing unit 76-1 together with a counter value it generates (hereinafter referred to as the clock counter value for checking).

[0067] The first image data reduction processing unit 75-1 reduces the amount of binary image data supplied from the first object detection processing unit 73-1, and supplies the resulting reduced image data to the first information transmission processing unit 76-1. This can shorten the processing time of the downstream fail-safe computer 33. Details of the processing of the first image data reduction processing unit 75-1 when the first object detection processing unit 73-1 uses the background difference method will be described later with reference to FIG. 6.

[0068] The first information transmission processing unit 76-1 supplies to the information processing unit 51-n of the fail-safe computer 33 the frame counter value extracted by the first image data acquisition unit 71-1, the first CRC value, the 2a CRC value supplied from the first CRC calculation processing unit 72-1, the image processing result supplied from the first object detection processing unit 73-1, the reduced image data supplied from the first image data reduction processing unit 75-1, the check clock counter value by the first clock check counter 74-1, and the extracted clock counter value.

[0069] 5 shows an example of the configuration of the functions of the first object detection processing unit 73-1 when the background subtraction method is used. The object detection processing unit is configured to include the functions of a background information storage unit 81-1 and a background subtraction method execution processing unit 82-1.

[0070] The background information storage unit 81-1 stores image information of a state where no object exists in the imaging range of the imaging unit 22, that is, the background. The background difference execution processing unit 82-1 compares the image information of the background stored in the background information storage unit 81-1 with the frame image data supplied from the first image data acquisition unit 71-1 for pixels other than the pixel values ​​described in the frame counter value, clock counter value, and CRC calculation value described with reference to FIG. 3, and detects that the pixel is an image of the background if the difference in pixel value is equal to or less than a threshold value, and detects that the pixel is an image of the object if the difference in pixel value is equal to or more than a threshold value. The background difference execution processing unit 82-1 supplies binary image data, in which pixels detected as having no object are set to 0 and pixels detected as having an object are set to 1, as processed image data to the first image data reduction processing unit 75-1, and supplies the image processing result (i.e., the detection result of the presence or absence of an object) to the first information transmission processing unit 76-1.

[0071] Here, suppose that an abnormality occurs due to some factor in the image information model of the background stored in the background information storage unit 81-1, and the range of pixel values ​​set as the background becomes wider. If an object is captured under such circumstances, the difference in pixel values ​​is sufficient in a normal system, so that it can be detected as an object, but in an abnormal system, the difference in pixel values ​​becomes small, so that it is determined as the background. This phenomenon becomes apparent when an object is captured in the imaging range, that is, when a detectable level of change in pixel values ​​occurs, causing a discrepancy in the processing results of the first image processing unit 41 and the second image processing unit 42. In contrast, while observing a state in which there is little change in the image, there is a possibility that the abnormality will continue to be latent.

[0072] However, in the image processing system 1, a test pattern is inserted at a frequency sufficiently high relative to the failure probability into the frame image data supplied from the imaging processing unit 11. If the test pattern has a sufficient pixel value difference from the background image, even if an abnormality occurs in the image information model of the background in either system, a mismatch in the processing results of the first image processing unit 41 and the second image processing unit 42 occurs at the timing when the test pattern is inserted, making it possible to detect the abnormality.

[0073] Next, the first image data reduction processing unit 75-1 executes the process of reducing the binary image data supplied from the first object detection processing unit 73-1 as described above. As described above, the fail-safe computer 33 has a redundant configuration, so safety is ensured, but the processing speed is relatively slow. That is, it is difficult for the fail-safe computer 33 to process images obtained from two or more imaging systems at a sufficient speed with the data capacity as it is. Therefore, the first image data reduction processing unit 75-1 reduces the binary image data supplied from the first object detection processing unit 73-1 so that the data to be compared is a size that can be collated by the fail-safe computer 33 and so as to allow for minute differences.

[0074] 6, the first image data reduction processing unit 75-1 divides the binary image data into a predetermined number of blocks, performs digesting to determine whether or not the value of "1" exists in each block, and supplies the result to the first information transmission processing unit 76-1. This reduced image data is used for a Hamming distance test in the Hamming distance test processing unit 103, which will be described later.

[0075] The first image data reduction processing unit 75-1 may divide the supplied binary image data into a predetermined number of blocks on average regardless of image position, or may change the weighting of the block division between the central part of the image, which is the main detection target in image processing, and an image position that is almost the same as the background image in the above-mentioned background subtraction method, in other words, the edge part that is not the detection target. This makes it possible to efficiently reduce image data while maintaining the accuracy of the calculation result of the Hamming distance described later.

[0076] 7 is a functional block diagram for explaining the functions of the information processing unit 51-n of the fail-safe computer 33. The information processing unit 51-n includes the functions of an information acquiring unit 91, a CRC information acquiring unit 92, a test pattern CRC information storage unit 93, a CRC information comparison processing unit 94, an image processing result comparing unit 95, and an output processing unit 96.

[0077] All the information processing units 51-n, which are redundant functions in the fail-safe computer 33, have similar functions, so illustrations and detailed descriptions thereof will be omitted here.

[0078] The information acquisition unit 91 acquires the 2a CRC value, the image processing result (i.e., the detection result of the presence or absence of an object), the reduced image data, the frame counter value, the check clock counter value, the extracted clock counter value, and the first CRC value supplied from the first information transmission processing unit 76-1 of the first image processing unit 41, and acquires the 2b CRC value, the image processing result, the reduced image data, the frame counter value, the check clock counter value, the extracted clock counter value, and the first CRC value supplied from the second information transmission processing unit 76-2 of the second image processing unit 42. Here, the information supplied from the first information transmission processing unit 76-1 of the first image processing unit 41 described above is referred to as the image processing result of the A system, etc., and the information supplied from the second information transmission processing unit 76-2 of the second image processing unit 42 described above is referred to as the image processing result of the B system, etc.

[0079] The CRC information acquisition unit 92 acquires the 2a CRC value, the 2b CRC value, and the first CRC value from the information acquired by the information acquisition unit 91, and supplies them to a CRC information comparison processing unit 94.

[0080] The test pattern CRC information storage unit 93 pre-stores a value corresponding to the first CRC value calculated based on information other than the pixel positions in which information other than image data described using Figure 3 is embedded in the test pattern generated by the test pattern generation unit 62, i.e., the correct CRC value when no fault occurs anywhere.

[0081] If the frame image data acquired by the information acquisition unit 91 is not a test pattern, the CRC information comparison processing unit 94 performs a comparison between the first CRC value and the 2a CRC value, a comparison between the first CRC value and the 2b CRC value, and a comparison between the 2a CRC value and the 2b CRC value. If these values ​​all match, the CRC information comparison processing unit 94 determines that no problem has occurred in the CRC test and supplies the comparison result to the output processing unit 96. If any of these values ​​do not match, the CRC information comparison processing unit 94 determines that a problem has occurred in the CRC test and supplies the comparison result to the output processing unit 96.

[0082] Here, if the CRC test is defective in both systems, it is detected that there is a failure in the image capture processing unit 11, and if the CRC test is defective in either system A or system B, it is detected that there is a failure in the system that output the defect, i.e., a failure in the first image processing unit 41 or the second image processing unit 42.

[0083] Furthermore, when the frame image data acquired by the information acquisition unit 91 is a test pattern, the CRC information comparison processing unit 94 performs the following operations: compares the first CRC value with the correct value stored in the test pattern CRC information storage unit 93; compares the 2a CRC value with the correct value stored in the test pattern CRC information storage unit 93; and compares the 2b CRC value with the correct value stored in the test pattern CRC information storage unit 93. When these values ​​all match, the CRC information comparison processing unit 94 determines that no problem has occurred in the CRC test and supplies the comparison result to the output processing unit 96. When any of these values ​​do not match, the CRC information comparison processing unit 94 determines that a problem has occurred in the CRC calculation of the mismatched portion and supplies the comparison result to the output processing unit 96.

[0084] The image processing result comparison unit 95 includes the functions of an image processing buffer 101 , a frame counter comparison processing unit 102 , a Hamming distance test processing unit 103 , a detection processing result comparison and determination unit 104 , and a clock counter comparison processing unit 105 .

[0085] The image processing buffer 101 holds the image processing results of system A, reduced image data, check clock counter value, extracted clock counter value, and frame counter value processed by the first image processing unit 41 and the second image processing unit 42, respectively, and the image processing results of system B, reduced image data, check clock counter value, extracted clock counter value, and frame counter value for two periods (the buffer capacity may be capable of holding more than two periods depending on the possibility of the range of period deviation that occurs), and detects the image processing results, reduced image data, check clock counter value, and extracted clock counter value to be checked based on the comparison result of the frame counter value by the frame counter comparison processing unit 102, and supplies the image processing results to the detection processing result comparison and judgment unit 104, the reduced image data to the Hamming distance test processing unit 103, and the check clock counter value and extracted clock counter value to the clock counter comparison processing unit 105, respectively.

[0086] The image processing unit 32 and the fail-safe computer 33 operate asynchronously even though they operate at the same cycle, and the first image processing unit 41 and the second image processing unit 42 operate asynchronously. Therefore, in order to correctly detect the match between the image processing performed in the first image processing unit 41 and the second image processing unit 42, an image processing buffer 101 capable of temporarily holding multiple frames of A-system and B-system data, and a frame counter comparison processing unit 102 for correctly selecting combinations to be compared from the image processing results held in the image processing buffer 101 are required.

[0087] If there are no image processing results for system A and system B with matching frame counter values ​​within the number of frames that can be held in the image processing buffer 101, the frame counter comparison processing unit 102 notifies the output processing unit 96 that a fault such as a dropped frame has occurred.

[0088] Detection of data to be checked will be described with reference to Figs. 8 and 9.

[0089] As shown in Figure 8, at a certain processing period t=1 of the fail-safe computer 33, if the frame counter values ​​acquired simultaneously in the image processing results of system A, etc. processed by the first image processing unit 41 and the second image processing unit 42 respectively stored in the image processing buffer 101 and the image processing results of system B, etc. match, then these image processing results, etc. become the objects of comparison, and the corresponding reduced images are supplied to the Hamming distance test processing unit 103, the image processing results are supplied to the detection processing result comparison judgment unit 104, and the check clock counter value and the extracted clock counter value are supplied to the clock counter comparison processing unit 105.

[0090] 9, in a certain processing cycle t=1 of the fail-safe computer 33, if the image processing results of the A system processed by the first image processing unit 41 and the second image processing unit 42 stored in the image processing buffer 101 do not match the frame counter value acquired simultaneously with the image processing results of the B system, the image processing results of the A system of two frames stored in the image processing buffer 101 are compared with the frame counter value acquired simultaneously with the image processing results of the B system in the next processing cycle t=2, and the image processing results to be checked are detected. The reduced image detected as the object of checking is supplied to the Hamming distance test processing unit 103, the image processing results are supplied to the detection processing result comparison and judgment unit 104, and the check clock counter value and the extracted clock counter value are supplied to the clock counter comparison processing unit 105.

[0091] The Hamming distance test processing unit 103 calculates the Hamming distance of each digest result of the reduced image of the A system and the reduced image of the B system of the same frame supplied from the image processing buffer 101. Specifically, the Hamming distance test processing unit 103 rearranges the digested binary data into a bit string as shown in Fig. 10, calculates the Hamming distance by comparing the bit string derived from the reduced image of the A system with the bit string derived from the reduced image of the B system as shown in Fig. 11, compares it with a predetermined threshold, and if the Hamming distance is smaller than the threshold, determines that there was no error in the Hamming test, and if the Hamming distance is larger than the threshold, determines that an error has occurred in the Hamming test, and supplies the determination result to the output processing unit 96.

[0092] Returning to Figure 7, the detection processing result comparison and judgment unit 104 judges whether the image processing results of system A and system B of the same frame supplied from the image processing buffer 101 match or mismatch, and supplies the judgment result to the output processing unit 96.

[0093] The clock counter comparison processor 105 obtains the check clock counter value and the extracted clock counter value from the image processing buffer 101, and compares the check clock counter value of system A with the extracted clock counter value, and compares the check clock counter value of system B with the extracted clock counter value of system B. If the comparison results of both systems A and B do not match, the clock counter comparison processor 105 supplies an abnormality in the clock counter value to the output processor 96. If the comparison results of at least either system A or system B match, it is not considered to be abnormal.

[0094] The output processing unit 96, based on each CRC test result supplied from the CRC information comparison processing unit 94, the Hamming test result supplied from the Hamming distance test processing unit 103, the image processing result supplied from the detection processing result comparison and judgment unit 104, i.e., the judgment result of whether the object presence or absence detection result is a match or mismatch, and whether or not an abnormality has been detected in the clock counter value by the clock counter comparison processing unit 105, if an abnormality is detected in any of them, it assumes that an error has been detected and supplies the detection content of the abnormality to the processing result judgment unit 52, and if no abnormality is detected, it supplies to the processing result judgment unit 52 a message that no error has been detected, as well as the image processing results of systems A and B supplied from the image processing buffer 101.

[0095] As described above, in the information processing units 51-n, when the image processing results are the same and no fault is detected by the fault detection process, the processing result determination unit 52 determines that the image processing results (here, the detection results of whether or not an object exists in the crossing) by the first image processing unit 41 and the second image processing unit 42 of the image processing unit 32 are correct, generates a control signal for executing various processes based on the result, and supplies it to the processing result output control unit 53. In addition, in the case where the image processing results are not the same in all the processing systems, or when a fault is detected by the fault detection process, the processing result determination unit 52 controls various operations to the safe side, specifically, for example, generates a control signal for commanding a signal operation to prevent a train from entering the corresponding crossing, and supplies it to the processing result output control unit 53. The processing result output control unit 53 controls the output of the control signal supplied from the processing result determination unit 52 to a predetermined other device, such as a management control device that manages railway signals.

[0096] Next, the processing of the imaging processing unit 11 will be described with reference to the flowchart of FIG.

[0097] In step S1, the imaging section 22 performs imaging processing at a predetermined frame rate and supplies the data to the data processing section 23. The imaging data acquisition section 61 of the data processing section 23 acquires frame image data obtained by imaging and supplies the data to the correction etc. processing section 63. Note that the processing of step S1 is executed as needed while the system is operating, without waiting for the completion of the processing of each step described below, and subsequent processing is executed for each frame image data.

[0098] In step S2, the test pattern generating section 62 determines whether it is time to insert a test pattern.

[0099] If it is determined in step S2 that it is time to insert a test pattern, the test pattern generating unit 62 generates a test pattern and supplies it to the correction processing unit 63 in step S3.

[0100] If it is determined in step S2 that it is not the timing to insert a test pattern, or after the processing of step S3 is completed, in step S4, the correction processing unit 63 performs conventional image correction processing, such as processing to correct pixel-by-pixel variations in the frame image data supplied from the imaging data acquisition unit 61 or the test pattern supplied from the test pattern generation unit 62, and supplies the corrected image data (however, the test pattern is not changed by the correction processing by the correction processing unit 63) to the image data frame generation unit 67 and the CRC generation unit 64.

[0101] In step S 5 , the CRC generating unit 64 generates a CRC based on the frame image data supplied from the correction processing unit 63 .

[0102] In step S6, the image data frame generation unit 67 generates data to be sent by adding the frame counter value, the clock counter value, and the first CRC value to predetermined locations in the image data area of ​​the frame image data, as described using Figure 3, to the frame image data including the test pattern supplied from the correction etc. processing unit 63, and supplies the data to the transmitting side interface processing unit 24.

[0103] In step S 7 , the transmitting interface processor 24 adds an FCS to the transmission data supplied from the image data frame generator 67 and transmits it to the determination processing unit 12 .

[0104] By this processing, as explained using Figure 3, frame image data including a test pattern is generated with the frame counter value, clock counter value and first CRC value added to predetermined locations in the image data area of ​​the frame image data, and the data to be sent is supplied to the judgment processing unit 12, so that even if an abnormality occurs in the non-redundant imaging processing unit 11, it can be reliably detected.

[0105] Then, after the processing of the flowchart in Figure 12 is completed, the receiving side interface processing unit 31 of the judgment processing unit 12 acquires information including the captured image data supplied from the imaging processing unit 11, performs FCS testing, and, if there are no problems, supplies the captured image data supplied from the imaging processing unit 11 to each of the first image processing unit 41 and the second image processing unit 42 of the image processing unit 32.

[0106] Next, the processing of the first image processing unit 41 and the second image processing unit 42 will be described with reference to the flowchart of Fig. 13. Here, the description will be given using the components of the first image processing unit 41, but since the first image processing unit 41 and the second image processing unit 42 execute the same processing, the description thereof will be omitted.

[0107] In step S21, the first image data acquisition unit 71-1 of the first image processing unit 41 acquires frame image data including the frame counter value, clock counter value, and first CRC value information described with reference to Fig. 3 from the receiving interface processing unit 31, extracts the frame counter value and the first CRC, and supplies them to the first information transmission processing unit 76-1. The first image data acquisition unit 71-1 also supplies the frame image data including the frame counter value, clock counter value, and first CRC information to the first CRC calculation processing unit 72-1, the first object detection processing unit 73-1, and the first clock check counter 74-1.

[0108] In step S22, the first CRC calculation processing unit 72-1 calculates a CRC based on information other than the pixel positions in which information other than image data described using Figure 3 is embedded by processing of the image data frame generation unit 67 from the frame image data supplied from the first image data acquisition unit 71-1, and supplies the calculated CRC value 2a to the first information transmission processing unit 76-1.

[0109] In step S23, the first clock check counter 74-1 extracts the clock counter value described in the frame image data supplied from the first image data acquisition unit 71-1, and supplies it to the first information transmission processing unit 76-1 together with the check clock counter value it generated.

[0110] In step S24, the first object detection processing unit 73-1 detects abnormalities within the imaging range using the frame image data supplied from the first image data acquisition unit 71-1, specifically, for example, executes image processing to detect whether or not an object such as a person or a car is present within the railroad crossing, and supplies the image processing result (i.e., the detection result of the presence or absence of an object) to the first information transmission processing unit 76-1, and supplies the binary image data, which is the processed image data, to the first image data reduction processing unit 75-1. Here, the first object detection processing unit 73-1 uses a background difference method.

[0111] In step S25, the first image data reduction processing unit 75-1 performs reduction processing on the binary image data supplied from the first object detection processing unit 73-1 as described using Figure 6, and supplies the resulting reduced image data to the first information transmission processing unit 76-1.

[0112] In step S26, the first information transmission processing unit 76-1 supplies the frame counter value extracted by the first image data acquisition unit 71-1, the first CRC value, the 2a CRC value supplied from the first CRC calculation processing unit 72-1, the check clock counter value supplied from the first clock check counter 74-1, the extracted clock counter value, the image processing result supplied from the first object detection processing unit 73-1, and the reduced image data supplied from the first image data reduction processing unit 75-1 to the information processing unit 51-n of the fail-safe computer 33.

[0113] By this processing, in each of the first image processing unit 41 and the second image processing unit 42 having the same functions, in addition to the image processing for object detection and the reduction processing thereof similar to the conventional ones, a part of the fault detection processing is executed, so that the fail-safe computer 33, which has a relatively slow processing speed, can perform comparison with the fault probability at a sufficiently high frequency.

[0114] Next, the CRC check process of the information processing unit 51-n will be described with reference to the flowchart of Fig. 14. The information processing unit 51-n executes the same process.

[0115] In step S31, the information acquisition unit 91 acquires the frame counter value, the first CRC value, the 2a CRC value, the check clock counter value, the extracted clock counter value, the image processing result, and the reduced image data supplied from the first information transmission processing unit 76-1 of the first image processing unit 41, and also acquires the frame counter value, the first CRC value, the 2b CRC value, the check clock counter value, the extracted clock counter value, the image processing result, and the reduced image data supplied from the second information transmission processing unit 76-2 of the second image processing unit 42. Here, the image processing result supplied from the first information transmission processing unit 76-1 of the first image processing unit 41 described above is referred to as the image processing result of the A system, and the image processing result supplied from the second information transmission processing unit 76-2 of the second image processing unit 42 described above is referred to as the image processing result of the B system.

[0116] In step S32, the CRC information acquisition unit 92 acquires the 2a CRC value, the 2b CRC value, and the first CRC value from the information acquired by the information acquisition unit 91, and supplies them to the CRC information comparison processing unit 94.

[0117] In step S33, the CRC information comparison processing unit 94 determines whether or not what has been acquired is a test pattern.

[0118] If it is determined in step S33 that what has been acquired is a test pattern, in step S34, the CRC information comparison processing unit 94 reads the CRC information in the test pattern, i.e., the correct value, from the test pattern CRC information storage unit 93, and compares the first CRC value with the correct value stored in the test pattern CRC information storage unit 93, compares the 2a CRC value with the correct value stored in the test pattern CRC information storage unit 93, and compares the 2a CRC value with the correct value stored in the test pattern CRC information storage unit 93.

[0119] In step S35, the CRC information comparison processor 94 determines whether or not all these CRCs match.

[0120] If it is determined in step S35 that any of the CRCs do not match, the CRC information comparison processor 94 outputs the content of the error to the output processor 96 in step S36.

[0121] If it is determined in step S33 that the acquired CRC value is not a test pattern, in step S37, the CRC information comparison processing unit 94 compares the first CRC value with the 2a CRC value, compares the first CRC value with the 2b CRC value, and compares the 2a CRC value with the 2b CRC value.

[0122] In step S38, the CRC information comparison processor 94 determines whether or not all these CRCs match.

[0123] If it is determined in step S38 that any of the CRCs do not match, the CRC information comparison processor 94 outputs the content of the error to the output processor 96 in step S39.

[0124] If it is determined in steps S35 and S38 that all the CRCs match, then in step S40, the CRC information comparison processing unit 94 outputs to the output processing unit 96 that there is no problem with the CRC check.

[0125] By performing such processing, a CRC test can be performed, making it possible to detect image data sticking and image data abnormalities in the transmission path.

[0126] Next, the image processing result comparison process of the information processing unit 51-n and the processing result determination unit 52 will be described with reference to the flowchart of Fig. 15. The information processing unit 51-n executes the same process.

[0127] In step S51, the image processing buffer 101 receives the image processing results of the A and B systems, the reduced image data, the check clock counter value, the extracted clock counter value, and the frame counter value.

[0128] In step S52, the frame counter comparison processing unit 102 extracts information in which the frame counter values ​​match between the A-system image processing results etc. and the B-system image processing results etc. stored in the image processing buffer 101, and supplies the reduced image to the Hamming distance test processing unit 103 and the image processing result to the detection processing result comparison and judgment unit 104.

[0129] In step S53, the clock counter comparison processor 105 obtains the check clock counter value and the extracted clock counter value from the image processing buffer 101, and compares the check clock counter value of system A with the extracted clock counter value, and also compares the check clock counter value of system B with the extracted clock counter value of system B. If the comparison results of both systems A and B do not match, the clock counter comparison processor 105 supplies an abnormality in the clock counter value to the output processor 96. If the comparison results of at least either system A or system B match, it is not considered to be abnormal.

[0130] In step S54, the Hamming distance test processing unit 103 calculates the Hamming distance for the reduced image of system A and the reduced image of system B supplied from the image processing buffer 101, as described with reference to Figures 10 and 11.

[0131] In step S55, the Hamming distance test processing unit 103 compares the calculation result of the Hamming distance with a predetermined threshold value, and determines whether or not the Hamming distance is less than the predetermined threshold value.

[0132] If it is determined in step S55 that the Hamming distance is not smaller than the predetermined threshold value, the Hamming distance test processing unit 103 supplies a Hamming error to the output processing unit 96 in step S56.

[0133] If it is determined in step S55 that the Hamming distance is smaller than the predetermined threshold, in step S57, the Hamming distance test processing unit 103 supplies the output processing unit 96 with information that a Hamming error has not occurred.

[0134] In step S58, the detection processing result comparison and judgment unit 104 judges whether the image processing results of system A and system B of the same frame supplied from the image processing buffer 101 match or not, and supplies the judgment result to the output processing unit 96.

[0135] In step S59, if an abnormality is detected in any of the CRC test results supplied from the CRC information comparison processing unit 94, the Hamming test result supplied from the Hamming distance test processing unit 103, the comparison result of the clock counter values ​​by the clock counter comparison processing unit 105, and the judgment result of whether the image processing results are equal or equal supplied from the detection processing result comparison judgment unit 104, the output processing unit 96 assumes that an error has been detected and supplies the detected abnormality to the processing result judgment unit 52; if no abnormality is detected, the output processing unit 96 supplies to the processing result judgment unit 52 a message that no error has been detected, as well as the image processing results of systems A and B supplied from the image processing buffer 101.

[0136] By such processing, the Hamming distance of the reduced binary image data is calculated, so that even a fail-safe computer 33 with a relatively slow processing speed can perform matching at a sufficiently high frequency.

[0137] When no fault is found in either the CRC check process of the information processing unit 51-n described using the flowchart of FIG. 14 or the image processing result comparison process of the information processing unit 51-n described using the flowchart of FIG. 15, and when it is determined that the image processing results match in all the information processing units 51-n, the processing result determination unit 52 determines that the image processing results by the first image processing unit 41 and the second image processing unit 42 of the image processing unit 32 are correct, generates a control signal for executing various processes based on the result, and supplies it to the processing result output control unit 53. When an error occurs in either unit, the processing result determination unit 52 controls various operations to the safety side, specifically, generates a control signal for commanding a signal operation to prevent a train from entering the corresponding railroad crossing, and supplies it to the processing result output control unit 53. The processing result output control unit 53 controls the output of the control signal supplied from the processing result determination unit 52 to a predetermined other device, such as a management control device that manages railway signals.

[0138] In this manner, the image processing system 1 includes the imaging processing unit 11, the first image processing unit 41 that processes an image captured by the imaging processing unit 11, the second image processing unit 42 that performs the same processing as the first image processing unit 41, and the fail-safe computer 33 as an information processing unit that performs processing for managing the operations of the imaging processing unit 11, the first image processing unit 41, and the second image processing unit 42 in two or more systems. The image processing system 1 includes the imaging processing unit 11, the first image processing unit 41, and the second image processing unit 42, and the fail-safe computer 33 as an information processing unit that performs processing in two or more systems for managing the operations of the imaging processing unit 11, the first image processing unit 41, and the second image processing unit 42. The first image processing unit 41 includes a first image data acquisition unit 71-1 that acquires an image to which a first CRC has been assigned by the image data frame generation unit 67, a first CRC calculation processing unit 72-1 that calculates a second CRC for the image acquired by the first image data acquisition unit 71-1, a first object detection processing unit 73-1 that executes processing for the image acquired by the first image data acquisition unit 71-1, and a first CRC calculation processing unit 74-1 that calculates a second CRC for the image acquired by the first image data acquisition unit 71-1. The second image processing unit 42 includes a second image data acquisition unit 71-2 that acquires an image to which the first CRC has been added by the image data frame generation unit 67, a second CRC calculation processing unit 72-2 that calculates a third CRC for the image acquired by the second image data acquisition unit 71-2, and a first information transmission processing unit 76-1 that transmits the second CRC calculated by the first CRC calculation processing unit 72-1 and the processing result by the first object detection processing unit 73-1 to the fail-safe computer 33. and a first information transmission processing unit 76-1 that transmits to the fail-safe computer 33 the first CRC applied to the image by the image data frame generation unit 67, the second CRC calculated by the first CRC calculation processing unit 72-1, and the processing result by the second object detection processing unit 73-2. The fail-safe computer 33 receives the first CRC applied to the image by the image data frame generation unit 67, the second CRC calculated by the first CRC calculation processing unit 72-1, andThe third CRC calculated by the second CRC calculation processing unit 72-2 is compared in two or more systems, and an output processing unit 96 is provided which detects a failure in any of the imaging processing unit 11, the first image processing unit 41, or the second image processing unit 42 based on the comparison result. This makes it possible to reliably detect fixation of image data caused by a failure in an imaging element or A / D converter in the non-redundant imaging unit 22, a failure in a read circuit of the correction processing unit 63, etc.

[0139] The imaging processing unit 11 further includes a test pattern generating section 62 that controls the supply of the test pattern, the image data frame generating section 67 assigns a first CRC to the test pattern whose supply is controlled by the test pattern generating section 62, the first CRC calculation processing section 72-1 of the first image processing section 41 further calculates a second CRC for the test pattern, and the second CRC calculation processing section 72-2 of the second image processing section 42 further calculates a third CRC for the test pattern, and the fail-safe computer 33: The output processing unit 96 further includes a test pattern CRC information storage unit 93 which stores the correct value of the CRC calculation result for the test pattern whose supply is controlled by the test pattern generation unit 62, and the output processing unit 96 compares the first CRC assigned to the test pattern by the image data frame generation unit 67, the second CRC corresponding to the test pattern calculated by the first CRC calculation processing unit 72-1, and the third CRC corresponding to the test pattern calculated by the second CRC calculation processing unit 72-2 with the correct values ​​stored in the test pattern CRC information storage unit 93, and detects a failure in any of the imaging processing unit 11, the first image processing unit 41, or the second image processing unit 42 based on the comparison result, thereby making it possible to correctly detect a failure in the non-redundant imaging processing unit 11.

[0140] The first image processing unit 41 has a background information storage unit 81-1 that stores a first background model for detecting an object by background subtraction, and a background subtraction execution processing unit 82-1 that performs background subtraction on the image and test pattern supplied from the imaging processing unit 11 using the first background model stored in the background information storage unit 81-1. The second image processing unit 42 has a background information storage unit 81-2 that stores a second background model that is identical to the first background model and for detecting an object by background subtraction, and a background subtraction execution processing unit 82-2 that performs background subtraction on the image and test pattern supplied from the imaging processing unit 11 using the second background model stored in the background information storage unit 81-2. Therefore, even if an abnormality occurs in the image information model of the background due to some factor, it is possible to prevent the abnormality from continuing to lie latent in a state where there is little change in the image.

[0141] The image data frame generation section 67 of the imaging processing unit 11 is configured to add a first CRC, a clock counter value, and a frame counter value to predetermined locations within the frame data of the captured image, so that even if the data frame structure is determined by a standard in communication between the imaging processing unit 11 and the judgment processing unit 12, it is possible to create this information at a subsequent stage.

[0142] The first information transmission processing unit 76-1 of the first image processing unit 41 further transmits the frame counter value provided by the image data frame generating unit 67, and the second information transmission processing unit 76-2 of the second image processing unit 42 further transmits the frame counter value provided by the image data frame generating unit 67. The fail-safe computer 33 includes an image processing buffer 101 which holds multiple frames of the processing result by the first object detection processing unit 73-1 and the frame counter value transmitted by the first information transmission processing unit 76-1, as well as the processing result by the second image processing execution unit and the frame counter value transmitted by the second information transmission processing unit 76-2; The image processing unit 32 and the fail-safe computer 33 operate asynchronously even if they operate at the same cycle, and the first image processing unit 41 and the second image processing unit 42 operate asynchronously, but it is possible to correctly select the combination to be compared, by referring to the frame counter value assigned to the processing result by the first object detection processing unit 73-1 and the processing result by the second object detection processing unit 73-2, which are stored in the image processing buffer 101.

[0143] The above-mentioned technology can be applied to hardware such as a personal computer.

[0144] The above-mentioned series of processes can be executed by hardware or software. When the series of processes are executed by software, the programs constituting the software are installed from a program recording medium into a computer incorporated in dedicated hardware, or into, for example, a general-purpose personal computer capable of executing various functions by installing various programs.

[0145] In addition, the program executed by the computer may be a program in which processing is performed chronologically in the order described in this specification, or it may be a program in which processing is performed in parallel or at the required timing, such as when called.

[0146] Furthermore, the present invention is not limited to the above-described embodiment, and various modifications are possible without departing from the spirit and scope of the present invention. [Explanation of symbols]

[0147] 1...image processing system, 11...imaging processing unit, 12...judgment processing unit, 21...control unit, 22...imaging unit, 23...data processing unit, 24...transmission side interface processing unit, 31...reception side interface processing unit, 32...image processing unit, 33...fail-safe computer, 41...first image processing unit, 42...second image processing unit, 51-n...information processing unit, 52...processing result judgment unit, 53...processing result output control unit, 61...imaging data acquisition unit, 62...test pattern generation unit, 63...correction etc. processing unit, 64...CRC generation unit, 65...frame counter, 66...clock counter, 67...image data frame generation unit, 71-1~2...first and second image data acquisition units, 72-1~2...first and second CRC calculation processing units, 73-1~2...first and second object detection processing units, 74-1-2...first and second clock check counters, 75-1-2...first and second image data reduction processing units, 76-1-2...first and second information transmission processing units, 91...information acquisition unit, 92...CRC information acquisition unit, 93...test pattern CRC information storage unit, 94...CRC information comparison processing unit, 95...image processing result comparison unit, 96...output processing unit, 101...image processing buffer, 102...frame counter comparison processing unit, 103...Hamming distance test processing unit, 104...detection processing result comparison and judgment unit, 105...clock counter comparison processing unit

Claims

1. An information processing system including an imaging unit, a first image processing unit that processes an image captured by the imaging unit, a second image processing unit that executes the same processing as that of the first image processing unit, and an information processing unit that executes processing for managing operations of the imaging unit, the first image processing unit, and the second image processing unit in two or more systems, The imaging unit includes: a detection information adding processing unit that calculates a first CRC based on the captured image and adds the first CRC to frame data of the image; The first image processing unit is a first image acquisition unit that acquires the image to which the first CRC has been added by the detection information addition processing unit; a first CRC calculation processing unit that calculates a second CRC of the image acquired by the first image acquisition unit; a first image processing execution unit that executes processing on the image acquired by the first image acquisition unit; a first sending unit that sends the first CRC added to the image acquired by the first image acquiring unit, the second CRC calculated by the first CRC calculation processing unit, and a processing result by the first image processing executing unit to the information processing unit; having The second image processing unit is a second image acquisition unit that acquires the image to which the first CRC has been added by the detection information addition processing unit; a second CRC calculation processing unit that calculates a third CRC of the image acquired by the second image acquisition unit; a second image processing execution unit that executes processing on the image acquired by the second image acquisition unit; a second sending unit that sends the first CRC added to the image acquired by the second image acquiring unit, the third CRC calculated by the second CRC calculation processing unit, and a processing result by the second image processing executing unit to the information processing unit; having The information processing unit includes: a failure detection unit that compares, in two or more systems, the first CRC added to the image by the detection information addition processing unit, the second CRC calculated by the first CRC calculation processing unit, and the third CRC calculated by the second CRC calculation processing unit, and detects a failure in any of the imaging unit, the first image processing unit, or the second image processing unit based on a comparison result; An information processing system comprising:

2. 2. The information processing system according to claim 1, The imaging unit includes: a test pattern supply control unit for controlling supply of the test pattern; the detection information adding processing unit adds the first CRC to the test pattern, the supply of which is controlled by the test pattern supply control unit; the first CRC calculation processing unit of the first image processing unit further calculates the second CRC for the test pattern; the second CRC calculation processing unit of the second image processing unit further calculates the third CRC for the test pattern; The information processing unit includes: a CRC storage unit that stores a correct answer value of a CRC calculation result for the test pattern, the supply of which is controlled by the test pattern supply control unit; The failure detection unit compares the first CRC added to the test pattern by the detection information addition processing unit, the second CRC corresponding to the test pattern calculated by the first CRC calculation processing unit, and the third CRC corresponding to the test pattern calculated by the second CRC calculation processing unit with the correct value stored in the CRC storage unit, and detects a failure in any of the imaging unit, the first image processing unit, or the second image processing unit based on a comparison result. An information processing system comprising:

3. 3. The information processing system according to claim 2, The first image processing unit is a first storage unit configured to store a first background model for detecting an object by a background subtraction method; a first background subtraction calculation unit that performs a background subtraction method on the image and the test pattern supplied from the imaging unit by using the first background model stored in the first storage unit; having The second image processing unit is a second storage unit configured to store a second background model, the second background model being the same as the first background model, for detecting the object by a background subtraction method; a second background subtraction calculation unit that performs a background subtraction method on the image and the test pattern supplied from the imaging unit by using the second background model stored in the second storage unit; Possesses An information processing system comprising:

4. The information processing system according to any one of claims 1 to 3, The detection information adding processing unit of the imaging unit adds the first CRC to a predetermined position in the frame data of the captured image. An information processing system comprising:

5. The information processing system according to any one of claims 1 to 4, The detection information adding processing unit further adds a clock counter value to a predetermined position in the frame data of the captured image. An information processing system comprising:

6. The information processing system according to any one of claims 1 to 5, The detection information adding processing unit further adds a frame counter value to a predetermined position in the frame data of the captured image. An information processing system comprising:

7. 7. The information processing system according to claim 6, the first sending unit of the first image processing unit further sends the frame counter value added by the detection information adding processing unit; the second sending unit of the second image processing unit further sends the frame counter value added by the detection information adding processing unit; The information processing unit includes: an image processing buffer that holds a processing result by the first image processing execution unit, the frame counter value sent by the first sending unit, and a processing result by the second image processing execution unit, and the frame counter value sent by the second sending unit for a plurality of frames; a comparison unit that refers to the frame counter value stored in the image processing buffer and compares a processing result by the first image processing execution unit and a processing result by the second image processing execution unit that correspond to the image of the same frame; An information processing system further comprising:

8. An information processing method for an information processing system including an imaging unit, a first image processing unit that processes an image captured by the imaging unit, a second image processing unit that executes the same processing as that of the first image processing unit, and an information processing unit that executes processing for managing operations of the imaging unit, the first image processing unit, and the second image processing unit in two or more systems, comprising: a first CRC calculation step of calculating a first CRC based on the image captured by the imaging unit; a detection information adding step of adding the first CRC calculated by the processing of the first CRC calculation step to frame data of the image; a first image acquisition step of acquiring the image to which the first CRC is added in the first image processing unit; a second image acquisition step of acquiring the image to which the first CRC is added in the second image processing unit; a second CRC calculation step of calculating a second CRC of the image acquired by the processing of the first image acquisition step in the first image processing unit; a third CRC calculation step of calculating a third CRC of the image acquired by the processing of the second image acquisition step in the second image processing unit; a first image processing step of processing the image acquired by the processing of the first image acquisition step in the first image processing unit; a second image processing step in which the second image processing unit processes the image acquired by the second image acquisition step; a first sending step of sending, in the first image processing unit, the first CRC added to the image, the second CRC calculated by the processing in the first CRC calculation step, and a processing result by the processing in the first image processing step to the information processing unit; a second sending step of sending, in the second image processing unit, the first CRC added to the image, the third CRC calculated by the processing in the second CRC calculation step, and a processing result by the processing in the second image processing step to the information processing unit; a failure detection step of comparing, in two or more systems, the first CRC, the second CRC, and the third CRC transmitted by the processing in the first transmission step and the second transmission step, and detecting a failure in any of the imaging unit, the first image processing unit, or the second image processing unit based on a comparison result, in the information processing unit; 13. An information processing method comprising:

Images