IMAGE PROCESSING SYSTEM, ENCODING DEVICE, ENCODING METHOD, AND ENCODING PROGRAM

Abstract

The present invention enables generation of decoded data that can be used in a plurality of types of processing by AI. This image processing system includes an encoding unit which encodes, when a plurality of types of processing by AI are to be performed for decoded data generated by decoding image data that is subjected to encoding, the image data at a compression ratio which is determined so that predetermined accuracy can be obtained for any of the processing results.

Description

[Technical field]

[0001] The present invention relates to an image processing system, an encoding device, an encoding method, and an encoding program. [Background technology]

[0002] In general, when recording or transmitting image data, the data size is reduced by encoding, thereby reducing the recording or transmission costs.

[0003] On the other hand, when recording or transmitting image data for use in processing by AI (Artificial Intelligence), a conceivable method would be to encode the image data by increasing the compression rate of each area to the limit at which the AI ​​can properly process the decoded data (i.e., at the limit compression rate). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2021-013146 [Patent Document 2] International Publication No. 2020 / 162495 Summary of the Invention [Problem to be solved by the invention]

[0005] However, there may be cases where multiple AI processes are included, with each process having a different limiting compression rate. For example, suppose that the multiple processes include an AI-based object detection process and an AI-based distance measurement process for the detected object. In this case, if the data is encoded at a limiting compression rate suitable for the object detection process, it may not be possible to obtain decoded data with the image quality required for distance measurement. As a result, although the object can be detected from the decoded data, it may not be possible to perform appropriate distance measurement for the detected object.

[0006] In one aspect, the present invention aims to generate decoded data that can be used in multiple AI processes. [Means for solving the problem]

[0007] According to one aspect, an image processing system includes: The system has an encoding unit that encodes image data at a compression rate determined so that when multiple AI processes are performed on the decoded data generated by decoding the encoded image data, the image data is encoded at a compression rate determined so that a predetermined level of accuracy is obtained for the results of all of the processes. Effect of the Invention

[0008] It will be possible to generate decoded data that can be used for multiple AI processes. [Brief description of the drawings]

[0009] [Figure 1] FIG. 1 is a diagram illustrating an example of a system configuration of an image processing system. [Diagram 2] FIG. 2 is a diagram illustrating an example of a hardware configuration of the encoding device and the image analysis device. [Diagram 3] FIG. 3 is a diagram showing a specific example of the compression rate control process. [Figure 4] FIG. 4 is a first flowchart showing the flow of the compression rate control process. [Figure 5A] FIG. 5A is a first diagram showing an example of a result of control of the compression rate. [Figure 5B] FIG. 5B is a second diagram showing an example of a result of control of the compression rate. [Figure 6] FIG. 6 is a diagram illustrating an example of a system configuration of an image processing system in a specific phase. [Figure 7] FIG. 7 is a diagram showing an example of a result of processing by each processing unit. [Figure 8] FIG. 8 is a diagram illustrating an example of rule information. [Figure 9] FIG. 9 is a flowchart showing the flow of the rule information generating process. [Figure 10] FIG. 10 is a diagram illustrating an example of a system configuration of an image processing system in the encoding phase. [Figure 11] FIG. 11 is a second flowchart showing the flow of the compression rate control process. [Figure 12] FIG. 12 is a diagram showing an example of a result of switching the compression rate. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0010] Hereinafter, each embodiment will be described with reference to the accompanying drawings. In this specification and the drawings, components having substantially the same functional configurations are denoted by the same reference numerals, and redundant description will be omitted.

[0011] [First embodiment] <Image processing system configuration> First, a system configuration of an image processing system including an encoding device and an image analysis device according to a first embodiment will be described. Fig. 1 is a diagram showing an example of the system configuration of an image processing system. As shown in Fig. 1, an image processing system 100 includes an imaging device 110, an encoding device 120, and an image analysis device 130. In the image processing system 100, the encoding device 120 and the image analysis device 130 are communicably connected via a network (not shown).

[0012] The imaging device 110 captures images at a predetermined frame rate and transmits the video data to the encoding device 120 .

[0013] An encoding program is installed in the encoding device 120, and the encoding device 120 functions as an encoding unit 121 and a compression rate setting unit 122 by executing the encoding program.

[0014] The encoding unit 121 encodes image data of each frame included in the video data to generate encoded data. When generating encoded data, the encoding unit 121 encodes the image data using a compression rate map (a map indicating the compression rate of each region when the image data is encoded at a different compression rate for each region) set by the compression rate setting unit 122. The encoding unit 121 also transmits the generated encoded data to the image analysis device 130.

[0015] Every time the compression rate setting unit 122 obtains a compression rate map generated by the image analysis device 130, the compression rate setting unit 122 sets the obtained compression rate map in the encoding unit 121. This allows the compression rate of each area to be appropriately controlled when the encoding unit 121 encodes image data.

[0016] An image analysis program is installed in the image analysis device 130. The image analysis device 130 functions as a decoding unit 131, processing units 132_1 to 132_n, an output unit 133, an accuracy monitoring unit 134, and a compression ratio map generating unit 135 by executing the image analysis program.

[0017] The decoding unit 131 decodes the coded data transmitted from the coding device 120 and generates decoded data.

[0018] The processing units 132_1 to 132_n each perform AI processing in parallel on the decoded data. The n types (n is an integer equal to or greater than 2) of AI processing may include, for example, processing for identifying attributes of each of n types of different objects included in the decoded data (for example, processing for identifying attributes of pedestrians, processing for identifying attributes of vehicles, etc.).

[0019] The output unit 133 outputs the results of the processing performed by the processing units 132_1 to 132_n on the decoded data.

[0020] The accuracy monitoring unit 134 performs accuracy monitoring processing based on rule information 136. The rule information 136 prescribes a rule for determining a compression rate when each of the processing units 132_1 to 132_n performs AI processing in parallel on the decoded data. Specifically, in the case of this embodiment, the rule information 136 prescribes that "the compression rate is determined so that the processing by any AI can obtain a result with a predetermined accuracy or higher."

[0021] For this reason, the accuracy monitoring unit 134 monitors whether or not a predetermined accuracy is obtained for the result of each of the processes performed by the processing units 132_1 to 132_n on the decoded data. When the accuracy monitoring unit 134 determines that the predetermined accuracy is not obtained for the result of any of the processes performed by the processing units 132_1 to 132_n, it determines a new compression ratio so that the predetermined accuracy is obtained.

[0022] When a new compression ratio is determined by the accuracy monitoring unit 134, the compression ratio map generating unit 135 generates a compression ratio map based on the new compression ratio that has been determined. In this embodiment, the compression ratio map generating unit 135 generates a compression ratio map in which the new compression ratio that has been determined is stored in all areas of the image data. The compression ratio map generating unit 135 generates a compression ratio map every time a new compression ratio is determined, and transmits the generated compression ratio map to the encoding device 120.

[0023] In the image processing system 100 of FIG. 1, the units surrounded by dashed lines execute a compression rate control process to appropriately control the compression rate of each region while the encoding unit 121, the decoding unit 131, the processing units 132_1 to 132_n, and the output unit 133 are operating.

[0024] <Hardware configuration of the encoding device and image analysis device> Next, a description will be given of the hardware configuration of the encoding device 120 and the image analysis device 130. Fig. 2 is a diagram showing an example of the hardware configuration of the encoding device and the image analysis device.

[0025] 2A is a diagram showing an example of a hardware configuration of an encoding device 120. The encoding device 120 includes a processor 201, a memory 202, an auxiliary storage device 203, an I / F (Interface) device 204, a communication device 205, and a drive device 206. The hardware components of the encoding device 120 are connected to each other via a bus 207.

[0026] The processor 201 has various arithmetic devices such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), etc. The processor 201 reads various programs (for example, an encoding program, etc.) onto the memory 202 and executes them.

[0027] The memory 202 has a main storage device such as a Read Only Memory (ROM) and a Random Access Memory (RAM). The processor 201 and the memory 202 form a so-called computer, and the processor 201 executes various programs read onto the memory 202, causing the computer to realize various functions.

[0028] The auxiliary storage device 203 stores various programs and various data used when the processor 201 executes the various programs.

[0029] The I / F device 204 is a connection device that connects the imaging device 110, which is an example of an external device, with the encoding device 120.

[0030] The communication unit 205 is a communication device for communicating with the image analysis unit 130 via a network.

[0031] The drive device 206 is a device for setting the recording medium 210. The recording medium 210 here includes media that record information optically, electrically, or magnetically, such as a CD-ROM, a flexible disk, a magneto-optical disk, etc. The recording medium 210 may also include semiconductor memories that record information electrically, such as ROM, flash memory, etc.

[0032] The various programs to be installed in the auxiliary storage device 203 are installed, for example, by setting the distributed recording medium 210 in the drive device 206 and reading out the various programs recorded in the recording medium 210 by the drive device 206. Alternatively, the various programs to be installed in the auxiliary storage device 203 may be installed by being downloaded from a network via the communication device 205.

[0033] 2B is a diagram showing an example of a hardware configuration of the image analysis device 130. Note that the hardware configuration of the image analysis device 130 is generally the same as the hardware configuration of the encoding device 120, so the following description will focus on the differences from the encoding device 120.

[0034] The processor 221 reads, for example, an image analysis program or the like onto the memory 222 and executes it.

[0035] The I / F device 224 accepts operations for the image analysis device 130 via an operation device 231. The I / F device 224 also outputs the results of processing by the image analysis device 130 and displays them via a display device 232. The communication device 225 also communicates with the encoding device 120 via a network.

[0036] <Specific example of compression ratio control process> Next, a specific example of the compression rate control process executed by each unit surrounded by a dashed line in Fig. 1 (the accuracy monitoring unit 134, the compression rate map generating unit 135, and the compression rate setting unit 122) in the image processing system 100 will be described. Fig. 3 is a diagram showing a specific example of the compression rate control process. In the example of Fig. 3, for the sake of simplicity, there are two processing units (processing unit 132_1 and processing unit 132_2).

[0037] In Fig. 3, reference numeral 310 denotes image data of each frame included in the video data. The example in Fig. 3 indicates that the situation around the shooting position has changed from "Scene A" to "Scene B" over time.

[0038] In addition, in Fig. 3, reference numeral 320 indicates the compression ratio of the entire region at each time, with the horizontal axis indicating time and the vertical axis indicating the compression ratio. According to the example of Fig. 3, the compression ratio map generating unit 135 generates a compression ratio map including a new compression ratio twice in response to a change in the situation around the shooting position.

[0039] 3, reference numeral 330 denotes the accuracy of the result of the processing performed by the processing unit 132_1 on the decoded data, the horizontal axis denotes time, and the vertical axis denotes accuracy. Also, the dotted line denotes the allowable accuracy.

[0040] According to the example of reference numeral 330 in Fig. 3, the result of the processing by the processing unit 132_1 is equal to or higher than the allowable accuracy at time t1, but is equal to or lower than the allowable accuracy at time t2 after the situation around the shooting position changes from "Scene A" to "Scene B". Moreover, the result of the processing by the processing unit 132_1 greatly exceeds the allowable accuracy at time t3 after the compression rate map including the new compression rate is set for the first time. Furthermore, the result of the processing by the processing unit 132_1 approaches the allowable accuracy at time t4 after the compression rate map including the new compression rate is set for the second time.

[0041] 3, reference numeral 340 denotes the accuracy of the result of the processing performed by the processing unit 132_2 on the decoded data, where the horizontal axis denotes time and the vertical axis denotes accuracy. Also, the dotted line denotes the allowable accuracy.

[0042] According to the example of reference numeral 340 in Fig. 3, the result of the processing by the processing unit 132_2 was equal to or higher than the allowable accuracy at time t1 and time t2, but approached the allowable accuracy between time t2 and time t3 after the situation around the shooting position changed from "Scene A" to "Scene B". Moreover, the result of the processing by the processing unit 132_2 greatly exceeds the allowable accuracy at time t3 after the compression rate map including the new compression rate is set for the first time. Furthermore, the result of the processing by the processing unit 132_1 approaches the allowable accuracy at time t4 after the compression rate map including the new compression rate is set for the second time.

[0043] Here, the compression rate control process will be described with reference to symbols 310, 330, and 340. While "scene A" continues, a default compression rate is set in the encoding unit 121, and both the results of the processing by the processing unit 132_1 and the processing by the processing unit 132_2 exceed the allowable accuracy.

[0044] On the other hand, as the situation around the shooting position changes from "Scene A" to "Scene B," the default compression rate is no longer appropriate, and the accuracy of the processing results by processing unit 132_1 and the accuracy of the processing results by processing unit 132_2 decrease.

[0045] As shown by reference numeral 330, when the result of the processing by the processing unit 132_1 falls below the allowable accuracy, the accuracy monitoring unit 134 determines a compression rate lower than the current compression rate as a new compression rate, as shown by reference numeral 320. Furthermore, the compression rate map generating unit 135 generates a compression rate map in which the determined new compression rate is stored in the entire area of ​​the image data, and the compression rate setting unit 122 sets the generated compression rate map in the encoding unit 121. This improves the image quality of the decoded data, and both the result of the processing by the processing unit 132_1 and the result of the processing by the processing unit 132_2 greatly exceed the allowable accuracy.

[0046] As indicated by reference numerals 330 and 340, the results of the processing by the processing units 132_1 and 132_2 greatly exceed the allowable accuracy, so that the accuracy monitoring unit 134 determines a compression rate higher than the current compression rate as a new compression rate, as indicated by reference numeral 320. Furthermore, the compression rate map generating unit 135 generates a compression rate map in which the determined new compression rate is stored in the entire area of ​​the image data, and the compression rate setting unit 122 sets the generated compression rate map in the encoding unit 121. As a result, both the results of the processing by the processing unit 132_1 and the results of the processing by the processing unit 132_2 approach the allowable accuracy.

[0047] In this way, by monitoring the accuracy of the results of processing by each processing unit and appropriately controlling the compression rate of each region according to the monitored accuracy, the results of processing by any AI can be made to have a predetermined accuracy. In other words, according to this embodiment, it becomes possible to generate decoded data that can be used in multiple processes by AI.

[0048] The method of calculating the accuracy of the processing result by each processing unit is arbitrary. For example, when the processing unit performs processing to identify the attribute of the object, the accuracy of the identified attribute or the change in the accuracy may be calculated as the accuracy of the processing result.

[0049] Furthermore, when the processing unit performs processing to detect an object, for example, the detection frequency or the transition of the detection frequency of the object in a plurality of decoded data may be calculated as the accuracy of the processing result.

[0050] In addition, when the processing unit performs processing to measure the distance to an object, for example, the fluctuation or the change in the fluctuation of the distance to the object in each of the multiple decoded data may be calculated as the accuracy of the processing result.

[0051] 3, the accuracy monitoring unit 134 monitored the accuracy of the results of the processing by the processing units 132_1 and 132_2 for the entire region of the image data. The compression ratio map generating unit 135 generated a compression ratio map by storing the determined new compression ratio in the entire region of the image data.

[0052] However, the method of generating the compression ratio map is not limited to this. For example, the accuracy monitoring unit 134 may monitor the accuracy of the results of the processing by the processing units 132_1 and 132_2 for each compression ratio setting granularity (for example, for each coding block in coding of a moving image) that can be set by the coding unit 121.

[0053] This allows the compression ratio map generating unit 135 to generate a compression ratio map storing the compression ratios determined for each set granularity. In this case, a compression ratio map is generated in which the compression ratios determined based on the accuracy of the results of processing by the processing units 132_1 and 132_2 are stored in each area for each set granularity in the image data.

[0054] Such a generation method is effective, for example, when the accuracy of the results of processing by each of the processing units 132_1, 132_2, ..., 132_n differs depending on the region in the image data or the type of object in the image data, etc. Alternatively, it is effective when integrating position information of objects whose attributes have been identified by the processing units 132_1, 132_2, ..., 132_n executed in parallel, and storing different compression rates for each region for each set granularity, etc.

[0055] <Compression ratio control process flow> Next, the flow of the compression rate control process will be described with reference to Fig. 4, which is a first flowchart showing the flow of the compression rate control process.

[0056] In step S401, the accuracy monitoring unit 134 reads out the rule information 136 that is defined in advance.

[0057] In step S402, the compression rate map generation unit 135 generates a compression rate map in which default compression rates are stored in the entire area of ​​the image data, and the compression rate setting unit 122 sets the generated compression rate map in the encoding unit 121.

[0058] In step S403, the accuracy monitoring unit 134 acquires the results of the processing by the processing units 132_1 to 132_n, and calculates the accuracy of the results of each processing.

[0059] In step S404, the accuracy monitoring unit 134 determines whether or not there is a processing unit whose processing result is below the allowable accuracy. If it is determined in step S404 that there is a processing unit whose processing result is below the allowable accuracy (YES in step S404), the process proceeds to step S407.

[0060] In this case, in step S407, the accuracy monitoring unit 134 determines a new compression rate based on the difference between the accuracy of the processing result and the allowable accuracy so that the processing result (or all of the processing results if multiple results are below the allowable accuracy) is above the allowable accuracy.

[0061] On the other hand, if it is determined in step S404 that there is no processing unit whose processing result is below the allowable accuracy (NO in step S404), the process proceeds to step S405.

[0062] In step S405, the accuracy monitoring unit 134 identifies the minimum accuracy among the accuracy of the results of the processing by each of the processing units 132_1 to 132_n, and calculates the difference between the identified minimum accuracy and the allowable accuracy.

[0063] In step S406, the accuracy monitoring unit 134 determines whether the calculated difference is equal to or greater than a predetermined threshold (i.e., whether the identified minimum accuracy exceeds the allowable accuracy by a predetermined threshold). If it is determined in step S406 that the difference is less than the predetermined threshold (NO in step S406), the process proceeds to step S409.

[0064] On the other hand, if it is determined in step S406 that the difference is equal to or greater than the predetermined threshold (YES in step S406), the process proceeds to step S407.

[0065] In this case, in step S407, the accuracy monitoring unit 134 determines a new compression rate according to the difference between the minimum accuracy and the allowable accuracy so that the minimum accuracy approaches the allowable accuracy.

[0066] In step S408, the compression ratio map generation unit 135 generates a compression ratio map in which the determined new compression ratio is stored in the entire area of ​​the image data, and the compression ratio setting unit 122 sets the generated compression ratio map in the encoding unit 121.

[0067] In step S409, the accuracy monitoring unit 134 determines whether or not to end the compression rate control process. If it is determined in step S409 that the compression rate control process is to be continued (NO in step S409), the process returns to step S403.

[0068] On the other hand, if it is determined in step S409 that the compression rate control process is to be ended (YES in step S409), the compression rate control process is ended.

[0069] <Compression ratio control results> Next, the results of compression rate control by the compression rate control process (FIG. 4) will be described. FIG. 5A is a first diagram showing an example of the results of compression rate control. In FIG. 5A, decoded data 510-513 indicate decoded data processed by processing unit 132_1 at times t1-t4 in FIG. 3, respectively. Also in FIG. 5A, decoded data 520-523 indicate decoded data processed by processing unit 132_2 at times t1-t4 in FIG. 3, respectively.

[0070] In the decoded data 510-513 and 520-523, the difference in the density of the hatching indicates the difference in the image quality of the decoded data due to the difference in the compression rate. Specifically, the lighter the hatching, the lower the compression rate and the higher the image quality of the decoded data, and the darker the hatching, the higher the compression rate and the lower the image quality of the decoded data.

[0071] The example of FIG. 5A shows that at time t1 before the situation around the shooting position changes, the processing unit 132_1 can identify the attribute of the object α, and the processing unit 132_2 can identify the attribute of the object β.

[0072] Moreover, the example of FIG. 5A shows that the processing unit 132_1 is no longer able to identify the attribute of the object α at time t2 due to a change in the situation around the shooting position.

[0073] In addition, the example of Figure 5A shows that when processing unit 132_1 is no longer able to identify the attributes of object α, the compression rate is appropriately controlled, so that at times t3 and t4, processing unit 132_1 is again able to identify the attributes of object α.

[0074] On the other hand, Fig. 5B is a second diagram showing an example of the control result of the compression ratio. The example of Fig. 5B shows a case where the accuracy is monitored for each compression ratio setting granularity, and here, for the sake of simplicity, a case where the accuracy is monitored in each of two regions is shown.

[0075] 5B, the processing unit 132_1 processes the decoded data 531, so that the processing unit 132_1 identifies only the attribute of the object α located below. Also, the processing unit 132_2 processes the decoded data 541, so that the processing unit 132_2 identifies both the attribute of the object β located below and the attribute of the object β located above.

[0076] In this way, when the result of processing by the processing unit 132_1 in the upper region falls below the allowable accuracy, the compression rate of the upper region is appropriately controlled.

[0077] The example of Figure 5B shows that by processing unit 132_1 processing the decoded data 532 after the compression rate of the upper region has been appropriately controlled, processing unit 132_1 is able to identify both the attributes of object α located below and the attributes of object α located above.

[0078] Moreover, the example of FIG. 5B shows that by processing unit 132_2 processing decoded data 542 after the compression rate of the upper region has been appropriately controlled, processing unit 132_2 is able to identify both the attributes of object β located below and the attributes of object β located above.

[0079] As is clear from the above description, the image processing system 100 according to the first embodiment performs multiple AI processes in parallel on decoded data generated by decoding encoded image data. Also, the image processing system 100 according to the first embodiment determines a compression ratio for encoding image data so that the results of all of the multiple AI processes have an acceptable accuracy, and encodes the image data at the determined compression ratio.

[0080] In this way, the image processing system 100 according to the first embodiment monitors the accuracy of the results of multiple AI processes, and appropriately controls the compression rate of each region according to the monitored accuracy. As a result, according to the image processing system 100 according to the first embodiment, even if the image quality of the decoded data is reduced due to a change in the situation around the shooting position, the compression rate can be changed to improve the image quality, so that the results of all AI processes can achieve acceptable accuracy.

[0081] In other words, according to the first embodiment, it becomes possible to generate decoded data that can be used in multiple processes by AI.

[0082] [Second embodiment] In the above first embodiment, a case where a plurality of processes are performed in parallel by AI has been described, whereas in the second embodiment, a case where a plurality of processes are performed sequentially by AI will be described.

[0083] When multiple processes are performed sequentially by AI, each process (e.g., the x+1th process) uses the results of the previous process (e.g., the xth process), so if decoded data with a different image quality from that of the previous process is required, it is necessary to switch the compression rate. For this reason, the image processing system according to the second embodiment is equipped with a function for sequentially switching the compression rate, and in order to realize this function, the image processing system according to the second embodiment is A determination phase to determine how to switch between compression ratios in a sequential manner; an encoding phase for sequentially switching the compression rate according to the identified switching method and encoding; Therefore, in the second embodiment, the following description will be divided into an identification phase and an encoding phase. However, the description will be centered on the differences from the first embodiment.

[0084] <System configuration of image processing system (specific phase)> First, a system configuration of an image processing system in a specific phase, which includes an encoding device and an image analysis device according to the second embodiment, will be described. Fig. 6 is a diagram showing an example of the system configuration of an image processing system in a specific phase. As shown in Fig. 6, an image processing system 600 in a specific phase includes an encoding device 620 and an image analysis device 630.

[0085] An encoding program is installed in the encoding device 620, and the encoding device 620 functions as the encoding unit 121 and a compression rate setting unit 622 by executing the encoding program.

[0086] Of these, the encoding unit 121 has the same functions as the encoding unit 121 described in the first embodiment with reference to Fig. 1, and therefore a description thereof will be omitted here. However, in the identification phase, the encoding unit 121 encodes image data of each frame included in the identification video data to generate encoded data.

[0087] Each time the compression rate setting unit 622 acquires a compression rate map generated by the image analysis device 630, it sets the acquired compression rate map in the encoding unit 121. In the specific phase, the compression rate setting unit 622 sets a compression rate map with gradually decreasing compression rates in the encoding unit 121 in order to search for a compression rate that can provide the image quality required by each of multiple AI processes.

[0088] An image analysis program is installed in the image analysis device 630. By executing the image analysis program, the image analysis device 630 functions as a decoding unit 131, a first processing unit 632_1, a second processing unit 632_2, ..., an n-th processing unit 632_n, and an output unit 133. In addition, the image analysis device 630 functions as a processing result analysis unit 634, a compression ratio map generation unit 635, and a rule information generation unit 636.

[0089] Of these, the decoding unit 131 and the output unit 133 have already been described in the first embodiment with reference to FIG. 1, and therefore a description thereof will be omitted here.

[0090] The first processing unit 632_1, the second processing unit 632_2, ..., the n-th processing unit 632_n each sequentially performs AI processing on the decoded data. The first processing unit 632_1 performs an object detection process to detect a predetermined object from the decoded data, The second processing unit 632_2 performs a distance measurement process to measure the distance to the detected object, A third processing unit 632_3 performs a situation determination process to determine the situation of the object based on the measured distance. Refers to processing units, etc.

[0091] In addition, the first processing unit 632_1, the second processing unit 632_2, ..., the nth processing unit 632_n process each decoded data generated by the decoding unit 131 decoding each encoded data that has been encoded by gradually decreasing the compression rate in a specific phase.

[0092] Furthermore, the first processing unit 632_1, the second processing unit 632_2, . . . the n-th processing unit 632_n notify the processing result analysis unit 634 of the results of the processing performed on each decoded data in the specific phase.

[0093] The processing result analysis unit 634 acquires the results of processing performed on each decoded data from the first processing unit 632_1, the second processing unit 632_2, ..., the n-th processing unit 632_n in the specific phase. Furthermore, the processing result analysis unit 634 analyzes the acquired processing results in the specific phase, searches for a compression ratio that can provide each processing unit with the image quality required for each processing unit to output an appropriate processing result, and identifies a switching compression ratio.

[0094] The compression rate map generating unit 635 sequentially generates a compression rate map in which stepwise lowered compression rates are stored in all areas in a specific phase, and transmits the generated compression rate map to the encoding device 620.

[0095] The rule information generation unit 636 identifies the switching condition for switching the compression rate and the area for switching, based on the switching compression rate of each processing unit found by the processing result analysis unit 634. Then, the rule information generation unit 636 generates rule information that specifies the compression rate decision rule when AI processing is performed sequentially on the decoded data.

[0096] Specifically, the rule information generated by the rule information generating unit 636 includes, in addition to the decision rules defined in the rule information 136 described in the first embodiment, the following: - Conditions for switching compression ratios, The area where the compression ratio is changed when the conditions are met. The compression ratio to switch to when the condition is met (switching compression ratio), The method includes a decision rule that specifies:

[0097] Thus, in the image processing system 600 in the specific phase, the components enclosed by the dashed dotted lines perform a rule information generation process for generating rule information.

[0098] <Results of processing by each processing section (specific phase)> Next, the results of the processing by each processing unit in the specific phase will be described below. Fig. 7 is a diagram showing an example of the results of the processing by each processing unit.

[0099] In FIG. 7, decoded data 710 to 730 indicate decoded data generated by encoding and decoding image data of each frame included in the video data for identification at a compression rate gradually decreased.

[0100] 7, decoded data 710 indicates decoded data in which an object is detected by an object detection process by the first processing unit 632_1. Decoded data 720 indicates decoded data in which an object is detected by the first processing unit 632_1 and the distance to the object is measured by a distance measurement process by the second processing unit 632_2. Decoded data 730 indicates decoded data in which an object is detected by the first processing unit 632_1, the distance to the object is measured by the second processing unit 632_2, and the situation of the object is judged by a situation judgment process by the third processing unit 632_3.

[0101] 7, by lowering the compression rate to 1, the first processing unit 632_1 can properly detect the object from the decoded data 710. As a result, the processing result analysis unit 634 can specify the compression rate 1 as the switching compression rate of the first processing unit 632_1.

[0102] 7, by lowering the compression rate to compression rate 2, the second processing unit 632_2 can properly measure the distance to the object in the area where the object is detected in the decoded data 720. This allows the processing result analysis unit 634 to specify compression rate 2 as the switching compression rate of the second processing unit 632_2.

[0103] 7, by lowering the compression rate to compression rate 3, the third processing unit 632_3 can appropriately determine the state of the object based on the distance to the object in the area where the object is detected in the decoded data 730. This allows the processing result analysis unit 634 to specify compression rate 3 as the switching compression rate of the third processing unit 632_3.

[0104] <Rule information details> Next, details of the rule information generated in the specific phase will be described. Fig. 8 is a diagram showing an example of the rule information. Again, for the sake of simplicity, a case will be described in which the image analyzing device 130 has a first processing unit 632_1 to a third processing unit 632_3.

[0105] Of these, the code 810 is Image quality required when the first processing unit 632_1 performs processing, Image quality required when the second processing unit 632_2 performs processing, Image quality required when the third processing unit 632_3 performs processing, are different from each other, (Image quality required when first processing unit 632_1 performs processing)<(Image quality required when second processing unit 632_2 performs processing)<(Image quality required when third processing unit 632_3 performs processing) The switching compression ratio is Compression ratio 1> Compression ratio 2> Compression ratio 3 This indicates that it has been identified as such.

[0106] In this case, the rule information 811 includes, in addition to the decision rules defined in the rule information 136 described in the first embodiment, The compression ratio of 1 is set for the entire image data area by default. On the condition that the result of the processing by the first processing unit 632_1 has been output, the compression ratio of the area of ​​the image data corresponding to the result of the processing by the first processing unit is switched to a compression ratio of 2; On the condition that the result of the processing by the second processing unit 632_2 has been output, the compression ratio of the area of ​​the image data corresponding to the result of the processing by the second processing unit is switched to a compression ratio of 3; The system will include decision rules that specify:

[0107] Also, the code 820 is Image quality required when the first processing unit 632_1 performs processing, Image quality required when the second processing unit 632_2 performs processing, Image quality required when the third processing unit 632_3 performs processing, are different from each other, (Image quality required when the first processing unit 632_1 performs processing)<(Image quality required when the second processing unit 632_2 performs processing)>(Image quality required when the third processing unit 632_3 performs processing) The switching compression ratio is Compression ratio 4> Compression ratio 5< Compression ratio 6 This indicates that it has been identified as such.

[0108] In this case, the rule information 821 includes, in addition to the decision rules defined in the rule information 136 described in the first embodiment, The compression ratio of 4 is set as the default for all areas of image data. On the condition that the result of the processing by the first processing unit 632_1 has been output, the compression ratio of the area of ​​the image data corresponding to the result of the processing by the first processing unit is switched to a compression ratio of 5; The system will include decision rules that specify:

[0109] Also, the code 830 is Image quality required when the first processing unit 632_1 performs processing, Image quality required when the second processing unit 632_2 performs processing, Image quality required when the third processing unit 632_3 performs processing, are different from each other, (Image quality required when the first processing unit 632_1 performs processing)>(Image quality required when the second processing unit 632_2 performs processing)>(Image quality required when the third processing unit 632_3 performs processing) The switching compression ratio is Compression ratio 7<Compression ratio 8<Compression ratio 9 This indicates that it has been identified as such.

[0110] In this case, the rule information 831 includes, in addition to the decision rules defined in the rule information 136 described in the first embodiment, The compression ratio of 7 is set as the default for all areas of the image data. The system will include decision rules that specify:

[0111] <Rule information generation process flow> Next, a flow of the rule information generation process by the image processing system 600 in the specific phase will be described below. Fig. 9 is a flowchart showing the flow of the rule information generation process.

[0112] In step S901, the encoding device 120 obtains video data for identification.

[0113] In step S902, the processing result analysis unit 634 of the image analysis device 630 sets "1" to a counter m, which indicates the processing order of each processing unit.

[0114] In step S903, the compression rate map generation unit 635 of the image analysis device 630 transmits a compression rate map in which predetermined compression rates are stored for all areas of the image data to the encoding device 120. In addition, the compression rate setting unit 622 of the encoding device 120 sets the compression rate map in the encoding unit 121.

[0115] In step S904, the encoding unit 121 of the encoding device 120 generates encoded data by encoding the image data included in the identification video data using the set compression rate map, and transmits the encoded data to the image analysis device .

[0116] In step S905, the decoding unit 131 of the image analyzing device 130 decodes the encoded data to generate decoded data.

[0117] In step S906, the first to m-th processing units of the image analysis device 130 process the decoded data.

[0118] In step S907, the processing result analysis unit 634 of the image analysis device 130 judges whether or not the mth processing unit has output an appropriate processing result. If it is judged in step S907 that the appropriate processing result has not been output (NO in step S907), the process proceeds to step S908.

[0119] In step S908, the compression rate map generating unit 635 of the image analyzing device 130 generates a compression rate map storing compression rates lowered by a predetermined step size for all areas of the image data, and transmits the generated compression rate map to the encoding device 120. Furthermore, the compression rate setting unit 622 of the encoding device 120 sets the compression rate map in the encoding unit 121. Then, the process returns to step S904.

[0120] On the other hand, if it is determined in step S907 that appropriate processing results have been output (YES in step S907), the process proceeds to step S909.

[0121] In step S909, the processing result analysis unit 634 of the image analysis device 130 specifies the compression ratio at which the appropriate processing result is output as the switching compression ratio for the m-th processing unit.

[0122] In step S910, the processing result analysis unit 634 of the image analysis device 130 determines whether or not the switching compression ratio has been specified for all processing units. If it is determined in step S910 that there is a processing unit for which the switching compression ratio has not been specified (NO in step S910), the process proceeds to step S911.

[0123] In step S911, the processing result analysis unit 634 of the image analysis device 630 increments the counter m indicating the processing order, and the process returns to step S903.

[0124] On the other hand, if it is determined in step S910 that the switching compression ratios have been identified for all processing units (YES in step S910), the process proceeds to step S912.

[0125] In step S912, the rule information generating unit 636 of the image analyzing device 630 generates rule information based on the switching compression ratios of the respective processing units, and ends the rule information generating process.

[0126] <System configuration of image processing system (encoding phase)> Next, a system configuration of an image processing system in the encoding phase, which includes an encoding device and an image analysis device according to a second embodiment, will be described. Fig. 10 is a diagram showing an example of the system configuration of an image processing system in the encoding phase. As shown in Fig. 10, an image processing system 1000 includes an imaging device 110, an encoding device 620, and an image analysis device 630.

[0127] An encoding program is installed in the encoding device 620, and the encoding device 620 functions as the encoding unit 121 and a compression rate setting unit 622 by executing the encoding program.

[0128] Of these, the encoding unit 121 has the same functions as the encoding unit 121 described with reference to FIG. 1 in the first embodiment, and therefore a description thereof will be omitted here.

[0129] Every time the compression rate setting unit 622 obtains a compression rate map generated by the image analysis device 630, the compression rate setting unit 622 sets the obtained compression rate map in the encoding unit 121. Note that in the encoding phase, the compression rate setting unit 622 A compression ratio map in which switching compression ratios based on rule information (e.g., any one of rule information 811 to 831) are stored in each area, or A compression ratio map in which the newly determined compression ratio for each region is stored for the switching compression ratio; Obtain one of the following:

[0130] An image analysis program is installed in the image analysis device 630. By executing the image analysis program, the image analysis device 630 functions as a decoding unit 131, a first processing unit 632_1, a second processing unit 632_2, ..., an n-th processing unit 632_n, an output unit 133, an accuracy monitoring unit 1034, and a compression ratio map generating unit 635.

[0131] Among these, the decoding unit 131, the first processing unit 632_1, the second processing unit 632_2, . . . the n-th processing unit 632_n, the output unit 133, and the compression ratio map generating unit 635 have already been described, and therefore description thereof will be omitted here.

[0132] The accuracy monitoring unit 1034 performs accuracy monitoring processing based on rule information (for example, any one of the rule information 811 to 831). As described above, the rule information 811 to 831 includes, in addition to the decision rules defined in the rule information 136 described in the first embodiment, - Conditions for switching compression ratios, The area where the compression ratio is changed when the conditions are met. The compression ratio to switch to when the condition is met (switching compression ratio), Therefore, when the first processing unit 632_1 to the n-th processing unit 632_n process the decoded data, the accuracy monitoring unit 1034 refers to any one of the rule information 811 to 831 and determines the switching compression ratio based on the result of the processing by each processing unit.

[0133] Furthermore, the accuracy monitoring unit 1034 monitors whether or not a predetermined accuracy is obtained for the results of processing by each processing unit. If the accuracy monitoring unit 134 determines that the predetermined accuracy is not obtained for the results of processing by any processing unit, it determines a new compression rate so that the predetermined accuracy is obtained.

[0134] When the accuracy monitoring unit 1034 notifies the compression ratio map generating unit 635 of a switching compression ratio, the compression ratio map generating unit 635 generates a compression ratio map by storing the switching compression ratio in a corresponding area. Moreover, when the accuracy monitoring unit 1034 notifies the compression ratio map generating unit 635 of a newly determined compression ratio, the compression ratio map generating unit 635 generates a compression ratio map by storing the newly determined compression ratio in a corresponding area. Moreover, every time the compression ratio map generating unit 635 generates a compression ratio map, the compression ratio map generating unit 635 transmits the generated compression ratio map to the encoding device 620.

[0135] <Compression ratio control process flow> Next, the flow of the compression rate control process in the encoding phase will be described. Fig. 11 is a second flowchart showing the flow of the compression rate control process. The differences from the compression rate control process described with reference to Fig. 4 in the first embodiment are steps S1101 to S1103 and step S1104.

[0136] In step S1101, the accuracy monitoring unit 1034 judges whether or not the results of the processing by the first processing unit 632_1 to the n-th processing unit 632_n satisfy the switching condition defined in the rule information.

[0137] In step S1101, if it is determined that the switching condition is not satisfied (NO in step S1101), the process proceeds to step S1103.

[0138] On the other hand, if it is determined in step S1101 that the switching condition is satisfied (YES in step S1101), the process proceeds to step S1102.

[0139] In step S1102, the compression ratio map generating unit 635 generates a compression ratio map by storing the switching compression ratio in a corresponding area. The compression ratio setting unit 622 sets the generated compression ratio map in the encoding unit 121.

[0140] In step S1103, the accuracy monitoring unit 1034 acquires the results of the processing by the first processing unit 632_1 to the n-th processing unit 632_n, and calculates the accuracy of the result of each processing.

[0141] In step S1104, the compression ratio map generation unit 635 generates a compression ratio map by storing the newly determined compression ratio in the corresponding area, and the compression ratio setting unit 622 sets the generated compression ratio map in the encoding unit 121.

[0142] <Results of switching compression ratio> Next, a specific example of the result of switching the compression rate by the compression rate control process (FIG. 11) will be described. FIG. 12 is a diagram showing an example of the result of switching the compression rate. In FIG. 12, decoded data 1210 to 1240 indicate decoded data generated by encoding and decoding image data of each frame included in the video data.

[0143] In the decoded data 1210 to 1240, the difference in the density of the hatching indicates the difference in the image quality of the decoded data due to the difference in the compression rate. Specifically, the lighter the hatching, the lower the compression rate and the higher the image quality of the decoded data, and the darker the hatching, the higher the compression rate and the lower the image quality of the decoded data.

[0144] The example of Fig. 12 shows a state where multiple objects appear over time. As shown in Fig. 12, the first processing unit 632_1 performs object detection processing on the entire area of ​​the decoded data 1210. At this time, the image data is encoded at a compression rate of 1.

[0145] For the decoded data 1220, the first processing unit 632_1 performs object detection processing on the entire region to detect the object 1201. At this time, the image data is encoded at a compression rate of 1, but as the object 1201 is detected, the compression rate of the region corresponding to the region where the object 1201 is detected is switched to a compression rate of 2 lower than the compression rate 1.

[0146] The decoded data 1230 is the image data, The area corresponding to the area where the object 1201 is detected is encoded with a compression rate of 2, and The area other than the area corresponding to the area where the object 1201 is detected is encoded with a compression ratio of 1. This is decoded data generated by decoding the encoded data.

[0147] For the decoded data 1230, the first processing unit 632_1 performs object detection processing on the entire region to detect the object 1201 and the object 1211. Furthermore, for the region in which the object 1201 is detected, the second processing unit 632_2 performs distance measurement processing to measure the distance to the object. Furthermore, as the distance measurement processing is performed on the object 1201, the compression rate of the region corresponding to the region in which the object 1201 is detected is switched to compression rate 3, which is lower than compression rate 2. Furthermore, as the object 1211 is newly detected, the compression rate of the region corresponding to the region in which the object 1211 is detected is switched to compression rate 2, which is lower than compression rate 1.

[0148] The decoded data 1240 is the image data, The area corresponding to the area where the object 1201 is detected is encoded with a compression ratio of 3, and The area corresponding to the area where the object 1211 is detected is encoded with a compression rate of 2, and The areas other than the areas corresponding to the areas where the object 1201 and the object 1211 are detected are encoded with a compression ratio of 1. This is decoded data generated by decoding the encoded data.

[0149] For the decoded data 1240, the first processing unit 632_1 performs an object detection process on the entire area to detect the object 1201, the object 1211, and the object 1221. For the area where the object 1201 is detected, the second processing unit 632_2 performs a distance measurement process to measure the distance to the object, and then the third processing unit 632_3 performs a situation determination process to determine the situation of the object.

[0150] Furthermore, the second processing unit 632_2 performs distance measurement processing on the region where the object 1211 is detected, and measures the distance to the object. In addition, as a result of performing the distance measurement processing on the object 1211, the compression ratio of the region corresponding to the region where the object 1211 is detected is switched to compression ratio 3, which is lower than compression ratio 2.

[0151] Furthermore, as a result of the object 1221 being newly detected, the compression rate of the area corresponding to the area in which the object 1221 is detected is switched to compression rate 2, which is lower than compression rate 1.

[0152] In the example of FIG. 12, no reference is made to a process for determining new compression ratios 1 to 3 so that the results of processing by each processing unit are equal to or higher than the allowable accuracy. However, the new compression ratio 1 may be determined, for example, by detecting the object 1201 in the decoded data 1220 and then calculating the accuracy of the results of processing by the first processing unit 632_1. Similarly, the new compression ratio 2 may be determined, for example, by performing a distance measurement process on the object 1201 in the decoded data 1230 and then calculating the accuracy of the results of processing by the second processing unit 632_2. Similarly, the new compression ratio 3 may be determined, for example, by performing a situation judgment process on the object 1201 in the decoded data 1240 and then calculating the accuracy of the results of processing by the third processing unit 632_3.

[0153] As is clear from the above description, the image processing system 1000 according to the second embodiment sequentially performs a plurality of processes by AI on each area of ​​the decoded data generated by decoding the encoded image data. Also, when the image quality required for the x+1th process is higher than the image quality required for the xth process, the image processing system 1000 according to the second embodiment switches the compression rate of the area corresponding to the result of the xth process to the switching compression rate.

[0154] In this way, in the image processing system 1000 according to the second embodiment, when multiple processes are performed sequentially by AI, the image is coded by switching to a switching compression rate that can provide the image quality required by each process. As a result, according to the image processing system 1000 according to the second embodiment, when multiple processes are performed sequentially by AI, each process can be appropriately performed.

[0155] Moreover, the image processing system 1000 according to the second embodiment determines the compression ratio for each region when encoding image data so that the results of all of the multiple processes performed by AI can obtain acceptable accuracy. Moreover, the image processing system 1000 according to the second embodiment encodes image data at the determined compression ratio for each region.

[0156] In this way, the image processing system 1000 according to the second embodiment monitors the accuracy of the results of multiple AI processes, and appropriately controls the compression rate after switching between regions according to the monitored accuracy. As a result, according to the image processing system 1000 according to the second embodiment, even if the image quality of the decoded data is reduced due to a change in the situation around the shooting position, the compression rate can be changed to improve the image quality, so that the results of all AI processes can achieve acceptable accuracy.

[0157] In other words, according to the second embodiment, it becomes possible to generate decoded data that can be used in multiple processes by AI.

[0158] [Other embodiments] In each of the above embodiments, the encoding device 120, 620 and the image analysis device 130, 630 are separate devices, but the encoding device 120, 620 and the image analysis device 130, 630 may be integrated into one device.

[0159] Furthermore, in each of the above embodiments, the imaging device 110 and the encoding device 120, 620 are separate devices, but the imaging device 110 and the encoding device 120, 620 may be integrated into one device.

[0160] In addition, when the devices are integrated, the functions described in the above embodiments are realized by, for example, executing an encoding program including the image analysis program in the device.

[0161] In each of the above embodiments, some of the functions realized by the image analysis devices 130, 630 may be realized by the encoding devices 120, 620, and some of the functions realized by the encoding devices 120, 620 may be realized by the image analysis devices 130, 630.

[0162] Furthermore, in the above second embodiment, it has been described that different image processing systems are used for the identification phase and the encoding phase, but the identification phase and the encoding phase may be executed using the same image processing system.

[0163] In addition, in the above second embodiment, the switching compression ratio is identified by gradually lowering the compression ratio in the identification phase. However, the switching compression ratio may be identified by gradually increasing the compression ratio.

[0164] In the above second embodiment, a new compression ratio is determined after switching to the switching compression ratio in the encoding phase. However, if a new compression ratio has already been determined, the switching compression ratio may be changed to the already determined new compression ratio.

[0165] In addition, in the above second embodiment, the compression ratio is described as being determined for each area in which an object is detected, but similar to the above first embodiment, the compression ratio may be determined for each setting granularity of the compression ratio.

[0166] In addition, the method of determining a new compression ratio by the accuracy monitoring unit described in each of the above embodiments is merely an example. For example, the compression ratio may be determined based on information for determining the compression ratio obtained by analyzing the recognition state and recognition process by the AI ​​and information indicating the progress of the accuracy of the processing result.

[0167] Specifically, when the compression ratio corresponding to the processing unit that processes the predetermined region is determined, the compression ratio may be increased (or decreased) stepwise while observing the progress of the accuracy of the processing result by the processing unit. Alternatively, the compression ratio may be determined based on information for determining the compression ratio obtained by analyzing the processing state or process of the processing unit that processes the predetermined region.

[0168] Alternatively, the compression rate may be determined based on information for determining the compression rate obtained by analyzing the processing state and processing progress of a processing unit that processes a specified area, and on the progress of the accuracy of the results of processing by a processing unit other than the processing unit that determines the compression rate.

[0169] In addition, the AI ​​recognition processing described in each of the above embodiments may include, in addition to deep learning processing, analysis processing that obtains results based on analysis by a computer or the like.

[0170] Although not described in the above embodiments, some areas in the image data may include areas where all processing units perform processing and areas where only some of the processing units perform processing. For areas where only some of the processing units perform processing, the compression ratio of the area may be determined based on the results of processing by those processing units. Alternatively, the compression ratio of the area may be determined including the results of processing by processing units that do not perform processing.

[0171] The present invention is not limited to the configurations shown in the above embodiments, combinations with other elements, etc. These points can be changed without departing from the spirit of the present invention, and can be appropriately determined according to the application form. [Explanation of symbols]

[0172] 100: Image processing system 110: Imaging device 120: Encoding device 121: Encoding section 122: Compression ratio setting section 130: Image analysis device 131: Decoding section 132_1 to 132_n: Processing section 133: Output section 134: Accuracy Monitoring Department 135: Compression ratio map generator 136: Rule information 620: Encoding device 622: Compression ratio setting section 632_1: First processing section 632_2: Second processing section 632_3: 3rd Processing Section 632_n: nth processing section 634: Processing result analysis unit 635: Compression ratio map generator 636: Rule information generation unit 811~831: Rule information 1034: Accuracy Monitoring Department

Claims

1. An encoding unit that encodes the image data at a compression rate determined so that a predetermined accuracy is obtained for each of the results of a plurality of processes by AI when the decoded data generated by decoding the encoded image data is subjected to the plurality of processes by AI; An image processing system comprising:

2. The image processing system according to claim 1 , further comprising a monitoring unit that monitors results of multiple processes performed by the AI.

3. The image processing system of claim 2, wherein when multiple processes by the AI ​​are performed in parallel and the result of any one of the results of the multiple processes by the AI ​​is below an allowable accuracy, the monitoring unit determines a new compression rate so that the result of the process is above an allowable accuracy.

4. The image processing system of claim 3, wherein when multiple processes by the AI ​​are performed in parallel and the smallest accuracy among the accuracy results of the multiple processes by the AI ​​is equal to or greater than an allowable accuracy, the monitoring unit determines a new compression rate so as to approach the allowable accuracy.

5. 3. The image processing system of claim 2, wherein when multiple processes are performed sequentially by the AI ​​and a higher image quality is required for performing the x+1th process than for performing the xth process, the monitoring unit switches the compression rate of the area corresponding to the result of the xth process to a compression rate lower than the compression rate used when the xth process was performed.

6. The method includes rule information that specifies each of the switching compression ratios capable of providing image quality required by each of the plurality of processes by the AI, a condition for switching to each switching compression ratio, and each area for switching to each switching compression ratio, The image processing system according to claim 5 , wherein the monitoring unit switches the compression rate for each area based on the results of the multiple processes performed by the AI ​​and the rule information.

7. 7. The image processing system according to claim 6, wherein, when a result of any one of the multiple processing results by the AI ​​is below an allowable accuracy, the monitoring unit determines a new compression rate for the area where the processing is performed so that the result of the processing is at or above the allowable accuracy.

8. The image processing system according to claim 6, wherein, when the smallest accuracy among the accuracy of the results of the multiple processes by the AI ​​is equal to or greater than an allowable accuracy, the monitoring unit determines a new compression rate for the area where the process is performed so as to approach the allowable accuracy.

9. An encoding unit that encodes the image data at a compression rate determined so that a predetermined accuracy is obtained for each of the results of a plurality of processes by AI when the decoded data generated by decoding the encoded image data is subjected to the plurality of processes by AI; An encoding device having the following construction.

10. When multiple processes by AI are performed on decoded data generated by decoding the encoded image data, the image data is encoded at a compression rate determined so that a predetermined accuracy is obtained for the results of all the processes. A coding method for computer-implemented processing.

11. When multiple processes by AI are performed on decoded data generated by decoding the encoded image data, the image data is encoded at a compression rate determined so that a predetermined accuracy is obtained for the results of all the processes. A coded program for causing a computer to carry out a process.

Images