WO2013128822A1 - 解析処理システム - Google Patents
解析処理システム Download PDFInfo
- Publication number
- WO2013128822A1 WO2013128822A1 PCT/JP2013/000796 JP2013000796W WO2013128822A1 WO 2013128822 A1 WO2013128822 A1 WO 2013128822A1 JP 2013000796 W JP2013000796 W JP 2013000796W WO 2013128822 A1 WO2013128822 A1 WO 2013128822A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- analysis
- engine
- quality
- processing
- input
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/217—Validation; Performance evaluation; Active pattern learning techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/96—Management of image or video recognition tasks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
Definitions
- the present invention relates to an analysis processing system, and particularly relates to an analysis processing system configured by combining a plurality of analysis engines.
- analysis engines that analyze various data have been developed.
- various analysis engines such as an analysis engine that generates position information for tracing human flow lines from moving image data, an analysis engine that identifies a person from still image data, and an analysis engine that generates text data from audio data have been developed. Has been.
- an analysis processing system that can obtain various analysis processing results from input data by combining a plurality of analysis engines of the same type or different types has been developed. For example, video data input from a camera is processed in parallel or in series using a person extraction engine, a flow line extraction engine, a face extraction engine, a face matching engine, etc., and an analysis process such as determining a person with a predetermined behavior is performed. It can be applied to the system to do.
- an analysis processing system composed of a large number of analysis engines it is desirable that the accuracy of the analysis result is guaranteed to be a certain level or more, and the processing time is fast, such as the analysis result being output in real time. It is desirable that
- Patent Document 1 discloses an image processing system in which an image providing apparatus and an image processing apparatus are connected.
- the image providing apparatus and the image processing apparatus exchange information such as resolution, and the image providing apparatus transfers the requested resolution image to the image processing apparatus.
- a technique is effective when data to be processed is transmitted and received between one-to-one devices, such as providing an image from an image providing device to an image processing device. It is difficult to apply to a system that processes a plurality of devices. Therefore, there is still a problem that it is difficult to design an analysis processing system configured by a plurality of analysis engines so as to satisfy appropriate analysis requirements.
- an object of the present invention is to solve the above-described problem that it is difficult to design an analysis processing system including a plurality of analysis engines.
- An analysis processing system is: Analysis processing execution means for controlling the operation of an analysis processing system configured by combining a plurality of analysis engines each performing predetermined analysis processing and executing analysis processing; The quality characteristics of the processing target data input for each of the analysis engines are determined so as to satisfy the accuracy requirements of the analysis results set in advance, and the analysis processing system is configured based on the determined quality characteristics.
- Input quality adjusting means for adjusting the quality characteristics of the processing target data input to the analysis engine; With The analysis processing execution means controls the operation of the analysis processing system so as to input the processing target data having a quality corresponding to the quality characteristic adjusted by the input quality adjustment means to the predetermined analysis engine;
- the configuration is as follows.
- the program which is the other form of this invention is: An information processing apparatus that controls the operation of an analysis processing system configured by combining a plurality of analysis engines each performing predetermined analysis processing to execute analysis processing, The quality characteristics of the processing target data input for each of the analysis engines are determined so as to satisfy the accuracy requirements of the analysis results set in advance, and the analysis processing system is configured based on the determined quality characteristics.
- Input quality adjusting means for adjusting the quality characteristics of the processing target data input to the analysis engine; Analysis processing for inputting the processing target data having the quality corresponding to the quality characteristic adjusted by the input quality adjusting means to the predetermined analysis engine, and controlling the operation of the analysis processing system to execute the analysis processing Execution means; It is a program for realizing.
- an analysis processing method includes: An analysis processing method for executing an analysis process by controlling the operation of an analysis processing system configured by combining a plurality of analysis engines each performing a predetermined analysis process, The quality characteristics of the processing target data input for each of the analysis engines are determined so as to satisfy the accuracy requirements of the analysis results set in advance, and the analysis processing system is configured based on the determined quality characteristics. Adjusting the quality characteristics of the processing target data input to the analysis engine; The processing target data of quality corresponding to the adjusted quality characteristic is input to the predetermined analysis engine, and the analysis processing is executed by controlling the operation of the analysis processing system.
- the configuration is as follows.
- the present invention is configured as described above, so that it is easy to design an analysis processing system including a plurality of analysis engines.
- FIG. 1 It is a block diagram which shows the structure of the analysis processing system in Embodiment 1 of this invention. It is a figure which shows an example of the data memorize
- FIGS. 1 to 5 are diagrams for explaining the configuration of the analysis processing system in the present embodiment
- FIGS. 6 to 10 are diagrams for explaining the operation of the analysis processing system.
- the analysis processing system shown in FIG. 1 includes an analysis processing device 10 configured by one or a plurality of information processing devices, a plurality of analysis engines 20 that respectively perform predetermined analysis processing, and a plurality of cameras 30 that capture images. And a data acquisition unit 40 that acquires video captured by the camera 30.
- the camera 30 is for acquiring video data (stream data) that is analysis target data in the present embodiment. For this reason, the camera 30 is installed at a plurality of preset locations, captures images of the locations, and outputs the captured image data to the data acquisition unit 40.
- the analysis target data in the analysis processing system of the present invention is not limited to video data, and may be any data such as audio data and image data.
- the data acquisition unit 40 acquires video data shot and output by the camera 30 with a preset quality, that is, with a preset resolution and frame rate. For example, the data acquisition unit 40 acquires video data with a resolution of 640 ⁇ 480 [dpi] and a frame rate of 30 [fps].
- the analysis engine 20 performs analysis processing on video data that is analysis target data by controlling the operation of the analysis processing apparatus 10.
- the analysis engine 20 in the present embodiment includes a thumbnail generation engine that generates a thumbnail image from video data, a moving body detection engine that detects a person from the video data, a person tracking engine that tracks the movement trajectory of the detected person, A face extraction engine that extracts a face portion of a detected person, and a face matching engine that specifies a person by matching the extracted face portion with a previously registered face.
- the analysis engine 20 is not limited to performing the above-described analysis processing, and may perform other analysis processing.
- the analysis processing device 10 includes an analysis processing execution unit 11 and a resolution determination unit 12 that are constructed by incorporating a program into the equipped arithmetic device.
- the analysis processing device 10 includes a processing flow storage unit 15 and a characteristic information storage unit 16 in the storage device that is equipped.
- each configuration will be described in detail.
- the processing flow storage unit 15 stores an analysis processing flow representing an analysis order in which the plurality of analysis engines 20 described above are combined.
- an example of the analysis processing flow stored in this embodiment is shown in FIGS.
- the analysis processing flow in the present embodiment is first a still image of a predetermined size representing one scene of video data from the video data acquired by the data acquisition unit 40 by the thumbnail generation engine 21.
- a thumbnail image is generated, and a series of flows F1 for storing the generated thumbnail image in the thumbnail storage unit 26 is provided.
- the flow of analysis processing from the thumbnail generation engine 21 to the thumbnail storage unit 26, which is the series of flows F1 is performed as a serial flow portion (serial flow analysis engine unit) having a serial processing order. It is specified by the determination unit 12.
- the thumbnail storage unit 26 is a partial function of the thumbnail generation engine 21 or a partial function of the analysis processing execution unit 11.
- the moving object detection engine 22 detects moving objects such as a person based on preset criteria from the video data acquired by the data acquisition unit 40.
- the person detected by the moving object detection engine 22 is tracked by the person tracking engine 23, and when it is determined that the movement trajectory of the person is the same as a specific trajectory based on a preset reference, an alert notification is sent.
- An alert is sent to the supervisor or the like in the unit 27.
- the face extraction engine 24 extracts a face portion from a person detected by the moving object detection engine 22 based on a preset criterion, and further registers in advance with the face portion extracted by the face matching engine 25. Whether the face data matches or not is checked, and if it matches, the alert notification unit 28 notifies an alert to the administrator or the like.
- the alert notification units 27 and 28 are a part of functions of each analysis engine or a part of functions of the analysis processing execution unit 11.
- the analysis process from the moving object detection engine 22 to the person tracking engine 23 / alert notification unit 27 and the analysis process from the moving object detection engine 22 to the face extraction engine 24 / face matching engine 25 / alert notification unit 28 are performed.
- the flow F ⁇ b> 2 that is included in the resolution determination unit 12 is a serial flow part in which the processing order is in series as a whole, although the processing order is branched to the human tracking engine 23 and the face extraction engine 24 of the moving object detection engine 22. Identified.
- the flow F21 of the analysis process from the person tracking engine 23 to the alert notification unit 27, which is further on the lower side from the lower branch point from the moving object detection engine 22, and the face extraction engine 24 to the face matching engine 25 / alert notification unit is specified by the resolution determination unit 12 described later as a serial subflow part (subflow analysis engine unit) having a serial processing order.
- analysis processing flow stored in the processing flow storage unit 15 may be stored in the above-described graph expression, and as shown in FIG. 3, the analysis engine flow in which the processing target data acquired for each camera 30 flows. It may be composed of data representing the order.
- the characteristic information storage unit 16 includes engine characteristic information in which the quality characteristic of the processing target data input to the analysis engine is associated with the accuracy characteristic of the analysis engine, and engine resource characteristic information indicating the resource characteristic of the analysis engine. I remember it.
- the engine characteristic information includes, for each analysis engine, the quality characteristics indicating the quality of the analysis target data input to the analysis engine and the accuracy of the analysis result by the analysis engine when the analysis target data of the quality is input.
- the accuracy characteristics to be represented are associated with each other.
- FIG. 4 shows an example of engine characteristic information.
- resolution [dpi] and frame rate [fps] are set as input information representing quality characteristics of video data that is analysis target data input to the analysis engine.
- accuracy information that represents the accuracy characteristics of the analysis engine the accuracy rate that indicates how accurate the analysis result is, the reproducibility that is an index that indicates how much leakage is present in the analysis result, and the analysis result Likelihood, which is an index indicating how much is correct, is set.
- the moving body detection engine shown in the first row of FIG. 4 has a resolution of 640 ⁇ in order to obtain the accuracy of the analysis result with a precision of 0.4, a recall of 0.5, and a likelihood of 0.3. It is set that it is only necessary to input data to be analyzed with a quality of 480 [dpi] and a frame rate of 5 [fps].
- the likelihood can be expressed as follows, for example.
- A) P (A
- B b) and P (A
- A P (A
- B b)
- the engine resource characteristic information stores the amount of resources used for the analysis processing by the analysis engine for each analysis engine.
- FIG. 5 shows an example of engine resource characteristic information.
- the processing time, CPU time, and memory usage required for analyzing data to be processed in a predetermined quality and in a predetermined unit are set as resource amounts.
- the moving body detection engine to which the processing target data is input from the camera 1 shown in the first row of FIG. 5 has a processing time of 15 [msec / msg], a CPU time of 10 [msec / msg], and a memory of 670 [KB]. It is set that the amount of resources is necessary.
- the resolution determination unit 12 is input to each analysis engine based on the engine characteristic information and the engine resource characteristic information stored in the characteristic information storage unit 16 and the analysis processing flow stored in the processing flow storage unit 15. Determine and adjust the quality characteristics of the data to be processed.
- the quality of the processing target data to be determined and adjusted is the frame rate, the quality is simply expressed as a rate or resolution.
- the resolution determination unit 12 refers to the engine characteristic information and the engine resource characteristic information to satisfy each of the analysis engines 20 so as to satisfy the accuracy requirement and the resource amount requirement of the analysis result set and requested in advance.
- the quality characteristics of the input data to be processed are determined (Yes in step S1 in FIG. 6, step S2).
- the resolution determination unit 12 sets, for each analysis engine, the quality characteristic of the processing target data that can output the analysis result while satisfying the accuracy of the set accuracy requirement as the minimum rate that is the minimum quality.
- the resolution determination unit 12 can perform processing with such a resource amount (resource amount requirement) when an upper limit value such as an allowable CPU usage time or a memory amount is set as the resource amount requirement for each analysis engine. Set the quality characteristics of the data to be processed as the maximum rate.
- the resolution determination unit 12 also sets an optimal rate that is an optimal quality characteristic of the processing target data input to each analysis engine based on the determined minimum rate and maximum rate. For example, the value of the lowest rate may be applied as it is as the optimum rate, and a value between the lowest rate and the highest rate is set according to a preset criterion. As an example, when the accuracy of the analysis results improves despite the small increase in the amount of resources used, the quality characteristics of the processing target data that is input when the amount of resources is used is set as the optimal rate. . As described above, as shown in FIG. 7, the resolution determination unit 12 determines the quality characteristics including the optimum rate, the lowest rate, and the highest rate for each of the analysis engines 21 to 25.
- the resolution determination unit 12 reads the information of the analysis processing flow from the processing flow storage unit 15, and determines a serial flow portion including one or a group of analysis engines whose processing order is in a serial relationship based on a preset reference. It is specified (step S4 in FIG. 6). For example, when the analysis processing flow is as shown in FIG. 2 or FIG. 3, two serial flow portions indicated by dotted lines F1 and F2 in FIG. 2 are specified.
- the resolution determination unit 12 checks whether there is a branch in the serial flow portions F1 and F2 (No in step S4 in FIG. 6, step S5). If there is a branch (Yes in step S5 in FIG. 6), one or a group of analysis engines whose processing order is in a serial relationship based on a preset criterion further downstream (lower side) from the branch point.
- the serial subflow part consisting of For example, in the example of FIG. 2, two serial subflow portions indicated by dotted lines F21 and F22 are specified in the serial flow portion F2. If there is a further branch in the specified serial subflow part, the serial subflow part is further specified on the downstream side of the branch point.
- the resolution determination unit 12 adjusts the quality characteristics of the processing target data input to the analysis engine positioned at the highest level in the serial flow portion and the serial subflow portion. (Steps S6, S7, S8, S9 in FIG. 6).
- the optimum rate 10 fps which is the quality characteristic determined for the person tracking engine 23, is set as it is. (See FIG. 8).
- the face extraction engine 24 and the face matching engine 25 exist in the serial subflow portion F21, among the optimum rates 1 fps and 3 fps, which are quality characteristics determined for each of them, A maximum value of 3 fps is set (see FIG. 8). Thereby, the resolution
- serial flow portion F2 since the serial subflow F21 and the serial subflow portion F22 exist downstream of the serial flow portion F2, among the optimum rates 10 fps and 3 fps, which are quality characteristics determined for each of them, A maximum value of 10 fps is set (see FIG. 9).
- the optimum rate 1 fps of the thumbnail generation engine 21 is set (step S9 in FIG. 6) (see FIG. 9).
- the resolution determination unit 12 determines the quality characteristics of the processing target data input to the analysis engine positioned at the top of the serial flow portion and the serial subflow portion from the analysis engine positioned at the top. Is adjusted so as to be the maximum value of the quality characteristic determined for the analysis engine whose processing order is positioned on the lower side. Thereby, the quality characteristic of the processing target data in the entire analysis processing flow can be set (steps S10 and S11 in FIG. 6).
- the quality characteristics of the serial flow portion F2 described above are not limited to adjustment based on the quality characteristics of the serial subflow portions F21 and F22 as described above, but the quality determined for all the analysis engines in the serial flow portion F2. You may adjust based on a characteristic.
- the quality characteristic of the processing target data input to each analysis engine is determined based on the engine characteristic information and the engine resource characteristic information set for each analysis engine.
- the determination may be based on information alone.
- a minimum rate that satisfies a predetermined accuracy requirement as a minimum is determined as a quality characteristic of the processing target data.
- the frame rate of the video data that is the processing target data is determined as the quality characteristic of the processing target data input to each analysis engine.
- the resolution of the video data is determined as the quality characteristic. Also good.
- the quality characteristics of the processing target data are not limited to the frame rate and resolution.
- the analysis processing execution unit 11 (analysis processing execution means) performs control so as to execute analysis processing on video data input from the camera 30 in a processing procedure based on the analysis processing flow stored in the processing flow storage unit 15. To do. At this time, the quality characteristic of the video data that is the processing target data input to each analysis engine is controlled.
- the analysis processing execution unit 11 adjusts the quality of the analysis engine at the top of the serial flow portions F1 and F2 and the serial subflow portions F21 and F22 as described above. Control is performed so that processing target data of quality corresponding to the characteristic is input. For example, as shown in FIG. 10, video data with a quality of 30 fps is input from the camera 30 to the data acquisition unit 40, but the serial flow portion F1 located on the lower side thereof has the highest data. Control is performed so that processing target data having a quality of 1 fps adjusted for the thumbnail generation engine 21 located therein is input. Similarly, the serial flow portion F2 positioned on the lower side of the data acquisition unit 40 is controlled to input the processing target data of quality 10 fps adjusted for the moving body detection engine 22 positioned at the uppermost position.
- control is performed so that processing target data having a quality of 10 fps adjusted with respect to the person tracking engine 23 positioned at the highest level is input to the serial subflow portion F21 positioned on the lower side of the serial flow portion F2. Then, the serial subflow portion F22 located on the lower side of the serial flow portion F2 is controlled to input the processing target data of quality 3 fps adjusted for the face extraction engine 24 located at the top.
- the analysis engine to which the processing target data with the quality higher than the optimum rate is input may be analyzed using only the processing target data with the required quality.
- the optimum rate is 5 fps, but since 10 fps processing target data is input, the processing target data input so as to use only data corresponding to 5 fps may be thinned out.
- the processing target data of the quality required to obtain the analysis result satisfying the required accuracy requirement is input to each analysis engine. Therefore, the analysis result by each analysis engine can satisfy the accuracy requirement, and an analysis processing system that can obtain a necessary analysis result can be easily designed.
- FIG. 11 is a diagram for explaining the configuration of the analysis processing system in the present embodiment.
- the analysis processing system according to the present embodiment has substantially the same configuration as that of the first embodiment.
- the analysis processing system according to the present embodiment includes a characteristic information generation unit 13 as shown in FIG.
- the characteristic information generation unit 13 (engine characteristic information generation unit, engine resource characteristic information generation unit) newly generates engine characteristic information and engine resource characteristic information stored in the characteristic information storage unit 16 described in the first embodiment. And stored in the characteristic information storage unit 16. Further, when engine characteristic information and engine resource characteristic information are already stored in the characteristic information storage unit 16, these information are updated.
- the characteristic information generation unit 13 sends the quality characteristics of the processing target data input to each analysis engine 20 and the quality characteristics to the analysis processing execution unit 11 that controls the operation of the analysis engine 20. Is obtained by requesting the measurement of the accuracy characteristics of the analysis result by each analysis engine 20 when the processing target data is input. Then, for each analysis engine 20, the acquired quality characteristic and accuracy characteristic are associated to generate engine characteristic information, and the generated engine characteristic information is stored in the characteristic information storage unit 16 as new engine characteristic information.
- the characteristic information generation unit 13 uses the amount of resources used when the analysis processing execution unit 11 that controls the operation of the analysis engine 20 performs analysis processing on a predetermined unit of data by each analysis engine 20. Request and obtain measurement of certain resource characteristics. Then, for each analysis engine 20, engine resource characteristic information is generated from the acquired resource characteristic, and the generated engine resource characteristic information is stored in the characteristic information storage unit 16 as new engine resource characteristic information.
- the quality characteristic of the processing target data input to each analysis engine 20 is determined and adjust.
- the analysis processing system in the present embodiment has substantially the same configuration as that of the first embodiment disclosed in FIG. 1, but in this embodiment, the characteristic information storage unit 16 (resource allocation information storage means) further includes a resource. The difference is that the allocation information is stored.
- the resource allocation information is information indicating the ratio of the resource amount of the information processing apparatus allocated to each analysis engine 20, and is set in advance by the camera 30 that captures the type and importance of the analysis engine and the processing target data. For example, a specific analysis engine that processes data to be processed photographed from a specific camera 30 is set to allocate more resources such as CPU time and memory capacity.
- the resolution determination unit 12 (input quality adjustment means) in the present embodiment sets the resource amount requirement that is allowed to be used in each analysis engine 20 based on the stored resource amount allocation information, and the resource amount The quality characteristics of the processing target data to be input to each analysis engine are determined so as to satisfy the requirements.
- Analysis processing execution means 111 for controlling the operation of an analysis processing system 110 configured by combining a plurality of analysis engines 120 each performing predetermined analysis processing and executing analysis processing; The quality characteristics of the data to be processed to be input for each analysis engine 120 are determined so as to satisfy the accuracy requirements of the preset analysis results, and the analysis processing system 110 is configured based on the determined quality characteristics.
- Input quality adjusting means 112 for adjusting the quality characteristic of the processing target data input to the predetermined analysis engine 120; With The analysis processing execution unit 111 operates the analysis processing system 110 so as to input the processing target data having a quality corresponding to the quality characteristic adjusted by the input quality adjustment unit 120 to the predetermined analysis engine 120. To control the Analysis processing system 110.
- An analysis processing flow storage means for storing an analysis processing flow representing an analysis order by the plurality of analysis engines;
- the input quality adjustment unit is configured to apply the predetermined analysis engine to the predetermined analysis engine based on the quality characteristic of the processing target data input for each analysis engine determined by the input quality adjustment unit and the analysis processing flow. Adjusting the quality characteristics of the input data to be processed; Analysis processing system.
- the analysis processing system according to attachment 2, wherein The input quality adjustment unit is configured to set the quality characteristic of the processing target data to be input to the analysis engine determined for the other analysis engine determined in the analysis processing flow with respect to the other analysis engine whose processing order is positioned on the lower side of the predetermined analysis engine. Based on the quality characteristics of the processing target data input to the predetermined analysis engine, Analysis processing system.
- the input quality adjusting means identifies a serial flow analysis engine unit composed of one or a group of the analysis engines having a serial relationship based on a criterion in which a processing order is set based on the analysis processing flow. Based on the quality characteristics of the processing target data input for each analysis engine determined for each analysis engine constituting the flow analysis engine unit, the processing order is positioned at the top of the serial flow analysis engine unit Adjusting the quality characteristic of the processing target data input to the predetermined analysis engine as the predetermined analysis engine, Analysis processing system.
- the analysis processing system according to attachment 4 wherein The input quality adjustment means is configured to calculate a maximum value of the quality characteristics of the processing target data input for each analysis engine determined for each analysis engine constituting the serial flow analysis engine unit, as the predetermined analysis. Set as the quality characteristic of the engine, Analysis processing system.
- the analysis processing system according to appendix 4 or 5
- the input quality adjustment means identifies a location where the processing order branches in the serial flow analysis engine unit based on the analysis processing flow, and is positioned lower than the branch location and the processing order is preset.
- the subflow analysis engine unit consisting of one or a group of the analysis engines that are in a series relationship with each other is specified, and the input is made for each analysis engine determined for each of the analysis engines constituting the subflow analysis engine unit.
- the analysis processing execution unit includes the processing target data having a quality corresponding to the quality characteristic of the processing target data input to the uppermost analysis engine of the subflow analysis engine unit adjusted by the input quality adjustment unit.
- Analysis processing system To control the operation of the analysis processing system so as to be input to the uppermost analysis engine of the subflow analysis engine unit, Analysis processing system.
- Appendix 7 The analysis processing system according to any one of appendices 1 to 6, Engine characteristic information that stores engine characteristic information in which the quality characteristic of the processing target data input to the analysis engine and the accuracy characteristic of the analysis result of the analysis engine are set in advance in association with each analysis engine.
- a storage means The input quality adjusting means determines the quality characteristics of the processing target data respectively input for each analysis engine so as to satisfy a precision requirement of a preset analysis result based on the engine characteristic information. Analysis processing system.
- the analysis processing system according to appendix 7,
- the engine characteristic information storage means stores, for each analysis engine, engine resource characteristic information in which a resource characteristic indicating a resource amount used for analysis processing by the analysis engine is preset.
- the input quality adjusting means is input for each analysis engine so as to satisfy the accuracy requirement of the analysis result and the resource amount requirement of the analysis processing set in advance based on the engine characteristic information and the engine resource characteristic information. Determining the quality characteristics of the data to be processed; Analysis processing system.
- appendix 11 The analysis processing system according to appendix 8 or 10, Resource allocation information storage means for storing resource allocation information representing resource allocation to the analysis engine set in advance, The input quality adjustment means sets the resource amount requirement based on the resource allocation information, and determines the quality characteristic of the processing target data input for each of the analysis engines so as to satisfy the resource amount requirement. Analysis processing system.
- An information processing apparatus that controls the operation of an analysis processing system configured by combining a plurality of analysis engines each performing predetermined analysis processing to execute analysis processing, The quality characteristics of the processing target data input for each of the analysis engines are determined so as to satisfy the accuracy requirements of the analysis results set in advance, and the analysis processing system is configured based on the determined quality characteristics.
- Input quality adjusting means for adjusting the quality characteristics of the processing target data input to the analysis engine; Analysis processing for inputting the processing target data having the quality corresponding to the quality characteristic adjusted by the input quality adjusting means to the predetermined analysis engine, and controlling the operation of the analysis processing system to execute the analysis processing Execution means;
- the input quality adjusting unit is configured to analyze the quality characteristic of the processing target data input for each of the analysis engines determined by the input quality adjusting unit, and an analysis process representing an analysis order by the plurality of analysis engines stored in advance. And adjusting the quality characteristic of the processing target data input to the predetermined analysis engine based on the flow, program.
- the above-described program is stored in a storage device or recorded on a computer-readable recording medium.
- the recording medium is a portable medium such as a flexible disk, an optical disk, a magneto-optical disk, and a semiconductor memory.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Signal Processing (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Combined Controls Of Internal Combustion Engines (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
Description
それぞれ所定の解析処理を行う複数の解析エンジンを組み合わせて構成される解析処理システムの動作を制御して解析処理を実行する解析処理実行手段と、
予め設定された解析結果の精度要件を満たすよう前記解析エンジン毎にそれぞれ入力される処理対象データの品質特性を決定すると共に、当該決定された品質特性に基づいて前記解析処理システムを構成する所定の前記解析エンジンに入力される前記処理対象データの前記品質特性を調整する入力品質調整手段と、
を備え、
前記解析処理実行手段は、前記入力品質調整手段にて調整された前記品質特性に対応する品質の前記処理対象データを、前記所定の解析エンジンに入力するよう前記解析処理システムの動作を制御する、
という構成をとる。
それぞれ所定の解析処理を行う複数の解析エンジンを組み合わせて構成される解析処理システムの動作を制御して解析処理を実行する情報処理装置に、
予め設定された解析結果の精度要件を満たすよう前記解析エンジン毎にそれぞれ入力される処理対象データの品質特性を決定すると共に、当該決定された品質特性に基づいて前記解析処理システムを構成する所定の前記解析エンジンに入力される前記処理対象データの前記品質特性を調整する入力品質調整手段と、
前記入力品質調整手段にて調整された前記品質特性に対応する品質の前記処理対象データを、前記所定の解析エンジンに入力し、前記解析処理システムの動作を制御して解析処理を実行する解析処理実行手段と、
を実現させるためのプログラムである。
それぞれ所定の解析処理を行う複数の解析エンジンを組み合わせて構成される解析処理システムの動作を制御して解析処理を実行する解析処理方法であり、
予め設定された解析結果の精度要件を満たすよう前記解析エンジン毎にそれぞれ入力される処理対象データの品質特性を決定すると共に、当該決定された品質特性に基づいて前記解析処理システムを構成する所定の前記解析エンジンに入力される前記処理対象データの前記品質特性を調整し、
調整された前記品質特性に対応する品質の前記処理対象データを、前記所定の解析エンジンに入力し、前記解析処理システムの動作を制御して解析処理を実行する、
という構成をとる。
本発明の第1の実施形態を、図1乃至図10を参照して説明する。図1乃至図5は、本実施形態における解析処理システムの構成を説明するための図であり、図6乃至図10は、解析処理システムの動作を説明するための図である。
次に、本発明の第2の実施形態を、図11を参照して説明する。図11は、本実施形態における解析処理システムの構成を説明するための図である。
次に、本発明の第3の実施形態を説明する。本実施形態における解析処理システムは、図1に開示した実施形態1のものとほぼ同様の構成を備えているが、本実施形態ではさらに、特性情報記憶部16(資源割当情報記憶手段)が資源割当情報を記憶している点で異なる。
上記実施形態の一部又は全部は、以下の付記のようにも記載されうる。以下、本発明における解析処理システム(図12参照)、プログラム、解析処理方法の構成の概略を説明する。但し、本発明は、以下の構成に限定されない。
それぞれ所定の解析処理を行う複数の解析エンジン120を組み合わせて構成される解析処理システム110の動作を制御して解析処理を実行する解析処理実行手段111と、
予め設定された解析結果の精度要件を満たすよう前記解析エンジン120毎にそれぞれ入力される処理対象データの品質特性を決定すると共に、当該決定された品質特性に基づいて前記解析処理システム110を構成する所定の前記解析エンジン120に入力される前記処理対象データの前記品質特性を調整する入力品質調整手段112と、
を備え、
前記解析処理実行手段111は、前記入力品質調整手段120にて調整された前記品質特性に対応する品質の前記処理対象データを、前記所定の解析エンジン120に入力するよう前記解析処理システム110の動作を制御する、
解析処理システム110。
付記1に記載の解析処理システムであって、
前記複数の解析エンジンによる解析順序を表す解析処理フローを記憶した解析処理フロー記憶手段を備え、
前記入力品質調整手段は、当該入力品質調整手段にて決定した前記解析エンジン毎に入力される前記処理対象データの前記品質特性と、前記解析処理フローと、に基づいて、前記所定の解析エンジンに入力される前記処理対象データの前記品質特性を調整する、
解析処理システム。
付記2に記載の解析処理システムであって、
前記入力品質調整手段は、前記解析処理フローにおいて前記所定の解析エンジンの下位側に処理順序が位置する他の前記解析エンジンについて決定した当該解析エンジンに入力される前記処理対象データの前記品質特性に基づいて、前記所定の解析エンジンに入力される前記処理対象データの前記品質特性を調整する、
解析処理システム。
付記3に記載の解析処理システムであって、
前記入力品質調整手段は、前記解析処理フローに基づいて、処理順序が予め設定された基準において直列の関係にある一つ又は一群の前記解析エンジンからなる直列フロー解析エンジン部を特定し、当該直列フロー解析エンジン部を構成する前記解析エンジン毎に決定した当該解析エンジン毎に入力される前記処理対象データの前記品質特性に基づいて、前記直列フロー解析エンジン部の最上位に処理順序が位置する前記解析エンジンを前記所定の解析エンジンとして当該所定の解析エンジンに入力される前記処理対象データの前記品質特性を調整する、
解析処理システム。
付記4に記載の解析処理システムであって、
前記入力品質調整手段は、前記直列フロー解析エンジン部を構成する前記解析エンジン毎について決定した当該解析エンジン毎に入力される前記処理対象データの前記品質特性のうちの最大値を、前記所定の解析エンジンの前記品質特性として設定する、
解析処理システム。
付記4又は5に記載の解析処理システムであって、
前記入力品質調整手段は、前記解析処理フローに基づいて、前記直列フロー解析エンジン部内で処理順序が分岐している箇所を特定し、当該分岐箇所よりも下位側に位置し処理順序が予め設定された基準において直列の関係にある一つ又は一群の前記解析エンジンからなるサブフロー解析エンジン部を特定し、当該サブフロー解析エンジン部を構成する前記解析エンジン毎に決定した当該解析エンジン毎に入力される前記処理対象データの前記品質特性に基づいて、前記サブフロー解析エンジン部の最上位に処理順序が位置する前記解析エンジンに入力される前記処理対象データの前記品質特性を調整し、
前記解析処理実行部は、前記入力品質調整手段にて調整された前記サブフロー解析エンジン部の最上位の前記解析エンジンに入力される前記処理対象データの前記品質特性に対応する品質の前記処理対象データを、当該サブフロー解析エンジン部の最上位の前記解析エンジンに入力するよう前記解析処理システムの動作を制御する、
解析処理システム。
付記1乃至6のいずれかに記載の解析処理システムであって、
前記解析エンジンに入力される処理対象データの品質特性と、前記解析エンジンの解析結果の精度特性と、が予め関連付けられて設定されたエンジン特性情報を、前記解析エンジン毎にそれぞれ記憶するエンジン特性情報記憶手段を備え、
前記入力品質調整手段は、前記エンジン特性情報に基づいて、予め設定された解析結果の精度要件を満たすよう前記解析エンジン毎にそれぞれ入力される前記処理対象データの前記品質特性を決定する、
解析処理システム。
付記7に記載の解析処理システムであって、
前記エンジン特性情報記憶手段は、前記解析エンジンによる解析処理に利用される資源量を表す資源特性が予め設定されたエンジン資源特性情報を、前記解析エンジン毎にそれぞれ記憶しており、
前記入力品質調整手段は、前記エンジン特性情報及び前記エンジン資源特性情報に基づいて、予め設定された解析結果の精度要件及び解析処理の資源量要件を満たすよう前記解析エンジン毎にそれぞれ入力される前記処理対象データの前記品質特性を決定する、
解析処理システム。
付記7又は8に記載の解析処理システムであって、
前記解析処理実行手段から、当該解析処理実行手段にて動作を制御されている前記解析エンジンに入力される前記処理対象データの品質特性と、前記解析エンジンによる解析結果の精度特性と、を前記解析エンジン毎にそれぞれ計測して取得し、当該取得した前記品質特性と前記精度特性とを関連付けて前記エンジン特性情報を生成して前記エンジン特性情報記憶手段に記憶する、あるいは、前記エンジン特性情報記憶手段に記憶されている前記エンジン特性情報を更新する、エンジン特性情報生成手段を備えた、
解析処理システム。
付記8に記載の解析処理システムであって、
前記解析処理実行手段から、当該解析処理実行手段にて動作を制御されている前記解析エンジンによる解析処理時に利用される資源量を表す資源特性を、前記解析エンジン毎にそれぞれ計測して取得し、当該取得した前記資源特性からなる前記エンジン資源特性情報を生成して、前記エンジン特性情報記憶手段に記憶する、あるいは、前記エンジン特性情報記憶手段に記憶されている前記エンジン資源特性情報を更新する、エンジン資源特性情報生成手段を備えた、
解析処理システム。
付記8又は10に記載の解析処理システムであって、
予め設定された前記解析エンジンに対する資源の割り当てを表す資源割当情報を記憶した資源割当情報記憶手段を備え、
前記入力品質調整手段は、前記資源割当情報に基づいて前記資源量要件を設定し、当該資源量要件を満たすよう前記解析エンジン毎にそれぞれ入力される前記処理対象データの前記品質特性を決定する、
解析処理システム。
それぞれ所定の解析処理を行う複数の解析エンジンを組み合わせて構成される解析処理システムの動作を制御して解析処理を実行する情報処理装置に、
予め設定された解析結果の精度要件を満たすよう前記解析エンジン毎にそれぞれ入力される処理対象データの品質特性を決定すると共に、当該決定された品質特性に基づいて前記解析処理システムを構成する所定の前記解析エンジンに入力される前記処理対象データの前記品質特性を調整する入力品質調整手段と、
前記入力品質調整手段にて調整された前記品質特性に対応する品質の前記処理対象データを、前記所定の解析エンジンに入力し、前記解析処理システムの動作を制御して解析処理を実行する解析処理実行手段と、
を実現させるためのプログラム。
付記12に記載のプログラムであって、
前記入力品質調整手段は、当該入力品質調整手段にて決定した前記解析エンジン毎に入力される前記処理対象データの前記品質特性と、予め記憶された前記複数の解析エンジンによる解析順序を表す解析処理フローと、に基づいて、前記所定の解析エンジンに入力される前記処理対象データの前記品質特性を調整する、
プログラム。
それぞれ所定の解析処理を行う複数の解析エンジンを組み合わせて構成される解析処理システムの動作を制御して解析処理を実行する解析処理方法であり、
予め設定された解析結果の精度要件を満たすよう前記解析エンジン毎にそれぞれ入力される処理対象データの品質特性を決定すると共に、当該決定された品質特性に基づいて前記解析処理システムを構成する所定の前記解析エンジンに入力される前記処理対象データの前記品質特性を調整し、
調整された前記品質特性に対応する品質の前記処理対象データを、前記所定の解析エンジンに入力し、前記解析処理システムの動作を制御して解析処理を実行する、
解析処理方法。
付記14に記載の解析処理方法であって、
前記複数の解析エンジンによる解析順序を表す解析処理フローが記憶されており、
前記エンジン特性情報に基づいて決定された前記解析エンジン毎に入力される前記処理対象データの前記品質特性と、前記解析処理フローと、に基づいて、前記所定の解析エンジンに入力される前記処理対象データの前記品質特性を調整する、
解析処理方法。
11 解析処理実行部
12 分解能決定部
15 処理フロー記憶部
16 特性情報記憶部
20 解析エンジン
21 サムネイル生成エンジン
22 動体検出エンジン
23 人物追跡エンジン
24 顔抽出エンジン
25 顔照合エンジン
26 サムネイル保存部
27,28 アラート通知部
30 カメラ
40 データ取得部
110 解析処理システム
111 解析処理実行手段
112 入力品質調整手段
120 解析エンジン
F1,F2 直列フロー部分
F21,F22 直列サブフロー部分
Claims (15)
- それぞれ所定の解析処理を行う複数の解析エンジンを組み合わせて構成される解析処理システムの動作を制御して解析処理を実行する解析処理実行手段と、
予め設定された解析結果の精度要件を満たすよう前記解析エンジン毎にそれぞれ入力される処理対象データの品質特性を決定すると共に、当該決定された品質特性に基づいて前記解析処理システムを構成する所定の前記解析エンジンに入力される前記処理対象データの前記品質特性を調整する入力品質調整手段と、
を備え、
前記解析処理実行手段は、前記入力品質調整手段にて調整された前記品質特性に対応する品質の前記処理対象データを、前記所定の解析エンジンに入力するよう前記解析処理システムの動作を制御する、
解析処理システム。 - 請求項1に記載の解析処理システムであって、
前記複数の解析エンジンによる解析順序を表す解析処理フローを記憶した解析処理フロー記憶手段を備え、
前記入力品質調整手段は、当該入力品質調整手段にて決定した前記解析エンジン毎に入力される前記処理対象データの前記品質特性と、前記解析処理フローと、に基づいて、前記所定の解析エンジンに入力される前記処理対象データの前記品質特性を調整する、
解析処理システム。 - 請求項2に記載の解析処理システムであって、
前記入力品質調整手段は、前記解析処理フローにおいて前記所定の解析エンジンの下位側に処理順序が位置する他の前記解析エンジンについて決定した当該解析エンジンに入力される前記処理対象データの前記品質特性に基づいて、前記所定の解析エンジンに入力される前記処理対象データの前記品質特性を調整する、
解析処理システム。 - 請求項3に記載の解析処理システムであって、
前記入力品質調整手段は、前記解析処理フローに基づいて、処理順序が予め設定された基準において直列の関係にある一つ又は一群の前記解析エンジンからなる直列フロー解析エンジン部を特定し、当該直列フロー解析エンジン部を構成する前記解析エンジン毎に決定した当該解析エンジン毎に入力される前記処理対象データの前記品質特性に基づいて、前記直列フロー解析エンジン部の最上位に処理順序が位置する前記解析エンジンを前記所定の解析エンジンとして当該所定の解析エンジンに入力される前記処理対象データの前記品質特性を調整する、
解析処理システム。 - 請求項4に記載の解析処理システムであって、
前記入力品質調整手段は、前記直列フロー解析エンジン部を構成する前記解析エンジン毎について決定した当該解析エンジン毎に入力される前記処理対象データの前記品質特性のうちの最大値を、前記所定の解析エンジンの前記品質特性として設定する、
解析処理システム。 - 請求項4又は5に記載の解析処理システムであって、
前記入力品質調整手段は、前記解析処理フローに基づいて、前記直列フロー解析エンジン部内で処理順序が分岐している箇所を特定し、当該分岐箇所よりも下位側に位置し処理順序が予め設定された基準において直列の関係にある一つ又は一群の前記解析エンジンからなるサブフロー解析エンジン部を特定し、当該サブフロー解析エンジン部を構成する前記解析エンジン毎に決定した当該解析エンジン毎に入力される前記処理対象データの前記品質特性に基づいて、前記サブフロー解析エンジン部の最上位に処理順序が位置する前記解析エンジンに入力される前記処理対象データの前記品質特性を調整し、
前記解析処理実行部は、前記入力品質調整手段にて調整された前記サブフロー解析エンジン部の最上位の前記解析エンジンに入力される前記処理対象データの前記品質特性に対応する品質の前記処理対象データを、当該サブフロー解析エンジン部の最上位の前記解析エンジンに入力するよう前記解析処理システムの動作を制御する、
解析処理システム。 - 請求項1乃至6のいずれかに記載の解析処理システムであって、
前記解析エンジンに入力される処理対象データの品質特性と、前記解析エンジンの解析結果の精度特性と、が予め関連付けられて設定されたエンジン特性情報を、前記解析エンジン毎にそれぞれ記憶するエンジン特性情報記憶手段を備え、
前記入力品質調整手段は、前記エンジン特性情報に基づいて、予め設定された解析結果の精度要件を満たすよう前記解析エンジン毎にそれぞれ入力される前記処理対象データの前記品質特性を決定する、
解析処理システム。 - 請求項7に記載の解析処理システムであって、
前記エンジン特性情報記憶手段は、前記解析エンジンによる解析処理に利用される資源量を表す資源特性が予め設定されたエンジン資源特性情報を、前記解析エンジン毎にそれぞれ記憶しており、
前記入力品質調整手段は、前記エンジン特性情報及び前記エンジン資源特性情報に基づいて、予め設定された解析結果の精度要件及び解析処理の資源量要件を満たすよう前記解析エンジン毎にそれぞれ入力される前記処理対象データの前記品質特性を決定する、
解析処理システム。 - 請求項7又は8に記載の解析処理システムであって、
前記解析処理実行手段から、当該解析処理実行手段にて動作を制御されている前記解析エンジンに入力される前記処理対象データの品質特性と、前記解析エンジンによる解析結果の精度特性と、を前記解析エンジン毎にそれぞれ計測して取得し、当該取得した前記品質特性と前記精度特性とを関連付けて前記エンジン特性情報を生成して前記エンジン特性情報記憶手段に記憶する、あるいは、前記エンジン特性情報記憶手段に記憶されている前記エンジン特性情報を更新する、エンジン特性情報生成手段を備えた、
解析処理システム。 - 請求項8に記載の解析処理システムであって、
前記解析処理実行手段から、当該解析処理実行手段にて動作を制御されている前記解析エンジンによる解析処理時に利用される資源量を表す資源特性を、前記解析エンジン毎にそれぞれ計測して取得し、当該取得した前記資源特性からなる前記エンジン資源特性情報を生成して、前記エンジン特性情報記憶手段に記憶する、あるいは、前記エンジン特性情報記憶手段に記憶されている前記エンジン資源特性情報を更新する、エンジン資源特性情報生成手段を備えた、
解析処理システム。 - 請求項8又は10に記載の解析処理システムであって、
予め設定された前記解析エンジンに対する資源の割り当てを表す資源割当情報を記憶した資源割当情報記憶手段を備え、
前記入力品質調整手段は、前記資源割当情報に基づいて前記資源量要件を設定し、当該資源量要件を満たすよう前記解析エンジン毎にそれぞれ入力される前記処理対象データの前記品質特性を決定する、
解析処理システム。 - それぞれ所定の解析処理を行う複数の解析エンジンを組み合わせて構成される解析処理システムの動作を制御して解析処理を実行する情報処理装置に、
予め設定された解析結果の精度要件を満たすよう前記解析エンジン毎にそれぞれ入力される処理対象データの品質特性を決定すると共に、当該決定された品質特性に基づいて前記解析処理システムを構成する所定の前記解析エンジンに入力される前記処理対象データの前記品質特性を調整する入力品質調整手段と、
前記入力品質調整手段にて調整された前記品質特性に対応する品質の前記処理対象データを、前記所定の解析エンジンに入力し、前記解析処理システムの動作を制御して解析処理を実行する解析処理実行手段と、
を実現させるためのプログラム。 - 請求項12に記載のプログラムであって、
前記入力品質調整手段は、当該入力品質調整手段にて決定した前記解析エンジン毎に入力される前記処理対象データの前記品質特性と、予め記憶された前記複数の解析エンジンによる解析順序を表す解析処理フローと、に基づいて、前記所定の解析エンジンに入力される前記処理対象データの前記品質特性を調整する、
プログラム。 - それぞれ所定の解析処理を行う複数の解析エンジンを組み合わせて構成される解析処理システムの動作を制御して解析処理を実行する解析処理方法であり、
予め設定された解析結果の精度要件を満たすよう前記解析エンジン毎にそれぞれ入力される処理対象データの品質特性を決定すると共に、当該決定された品質特性に基づいて前記解析処理システムを構成する所定の前記解析エンジンに入力される前記処理対象データの前記品質特性を調整し、
調整された前記品質特性に対応する品質の前記処理対象データを、前記所定の解析エンジンに入力し、前記解析処理システムの動作を制御して解析処理を実行する、
解析処理方法。 - 請求項14に記載の解析処理方法であって、
前記複数の解析エンジンによる解析順序を表す解析処理フローが記憶されており、
前記エンジン特性情報に基づいて決定された前記解析エンジン毎に入力される前記処理対象データの前記品質特性と、前記解析処理フローと、に基づいて、前記所定の解析エンジンに入力される前記処理対象データの前記品質特性を調整する、
解析処理方法。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2014501999A JP6225896B2 (ja) | 2012-03-02 | 2013-02-14 | 解析処理システム |
US14/379,541 US20150030260A1 (en) | 2012-03-02 | 2013-02-14 | Analysis system |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2012046674 | 2012-03-02 | ||
JP2012-046674 | 2012-03-02 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2013128822A1 true WO2013128822A1 (ja) | 2013-09-06 |
Family
ID=49082038
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2013/000796 WO2013128822A1 (ja) | 2012-03-02 | 2013-02-14 | 解析処理システム |
Country Status (3)
Country | Link |
---|---|
US (1) | US20150030260A1 (ja) |
JP (1) | JP6225896B2 (ja) |
WO (1) | WO2013128822A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016028143A1 (en) * | 2014-08-20 | 2016-02-25 | Mimos Berhad | Method and apparatus for performing parallel video analytics |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP3142040A1 (en) * | 2015-09-11 | 2017-03-15 | Canon Kabushiki Kaisha | Information processing apparatus, method of controlling the same, and program |
JP6775343B2 (ja) * | 2015-09-11 | 2020-10-28 | キヤノン株式会社 | 情報処理装置及びその制御方法及びプログラム |
KR102669100B1 (ko) | 2018-11-02 | 2024-05-27 | 삼성전자주식회사 | 전자 장치 및 그 제어 방법 |
WO2020157885A1 (ja) * | 2019-01-31 | 2020-08-06 | 日本電気株式会社 | データストリーム割り当て方法、システムおよびプログラム |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH11136499A (ja) * | 1997-10-27 | 1999-05-21 | Canon Inc | 画像処理システム及び画像処理装置 |
JP2001175872A (ja) * | 1999-12-15 | 2001-06-29 | Clarion Co Ltd | 画像処理装置及び方法並びに画像処理用ソフトウェアを記録した記録媒体 |
JP2011103561A (ja) * | 2009-11-11 | 2011-05-26 | Sony Corp | 画像処理システム、画像処理装置、画像処理方法、およびプログラム |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030210271A1 (en) * | 2002-05-13 | 2003-11-13 | King William Davis | Power based level-of- detail management system for a portable computer graphics display |
KR100679046B1 (ko) * | 2005-10-10 | 2007-02-05 | 삼성전자주식회사 | 이미지의 지역적 정보를 이용한 비트 해상도 확장 방법 및장치 |
US20100199189A1 (en) * | 2006-03-12 | 2010-08-05 | Nice Systems, Ltd. | Apparatus and method for target oriented law enforcement interception and analysis |
US8411940B2 (en) * | 2007-08-15 | 2013-04-02 | I.R.I.S. | Method for fast up-scaling of color images and method for interpretation of digitally acquired documents |
JP5429445B2 (ja) * | 2008-04-08 | 2014-02-26 | 富士フイルム株式会社 | 画像処理システム、画像処理方法、およびプログラム |
US8805110B2 (en) * | 2008-08-19 | 2014-08-12 | Digimarc Corporation | Methods and systems for content processing |
JP5284537B2 (ja) * | 2010-03-31 | 2013-09-11 | キヤノン株式会社 | 画像処理装置、画像処理方法、画像処理プログラム、およびそれを用いた撮像装置 |
US20120287487A1 (en) * | 2011-05-11 | 2012-11-15 | Xerox Corporation | System and method for determining scan limits with control tiers and automatically scanning documents according to same |
-
2013
- 2013-02-14 WO PCT/JP2013/000796 patent/WO2013128822A1/ja active Application Filing
- 2013-02-14 JP JP2014501999A patent/JP6225896B2/ja active Active
- 2013-02-14 US US14/379,541 patent/US20150030260A1/en not_active Abandoned
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH11136499A (ja) * | 1997-10-27 | 1999-05-21 | Canon Inc | 画像処理システム及び画像処理装置 |
JP2001175872A (ja) * | 1999-12-15 | 2001-06-29 | Clarion Co Ltd | 画像処理装置及び方法並びに画像処理用ソフトウェアを記録した記録媒体 |
JP2011103561A (ja) * | 2009-11-11 | 2011-05-26 | Sony Corp | 画像処理システム、画像処理装置、画像処理方法、およびプログラム |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016028143A1 (en) * | 2014-08-20 | 2016-02-25 | Mimos Berhad | Method and apparatus for performing parallel video analytics |
Also Published As
Publication number | Publication date |
---|---|
JPWO2013128822A1 (ja) | 2015-07-30 |
JP6225896B2 (ja) | 2017-11-08 |
US20150030260A1 (en) | 2015-01-29 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6225896B2 (ja) | 解析処理システム | |
US10559062B2 (en) | Method for automatic facial impression transformation, recording medium and device for performing the method | |
CN109891189B (zh) | 策划的摄影测量 | |
US10257502B2 (en) | Methods and apparatus for controlling light field capture | |
US10198660B2 (en) | Method and apparatus for event sampling of dynamic vision sensor on image formation | |
US10269130B2 (en) | Methods and apparatus for control of light field capture object distance adjustment range via adjusting bending degree of sensor imaging zone | |
US9571726B2 (en) | Generating attention information from photos | |
CN103546692A (zh) | 一种实现一体化摄像机自动聚焦的方法及系统 | |
JP2015033103A (ja) | 画像処理装置、画像処理方法、およびプログラム | |
WO2021135639A1 (zh) | 活体检测方法及装置 | |
WO2023040423A1 (zh) | 镜头的调焦方法、装置、计算机设备和存储介质 | |
CN113066092A (zh) | 视频对象分割方法、装置及计算机设备 | |
WO2014091667A1 (ja) | 解析制御システム | |
CN105701762A (zh) | 一种图片处理方法和电子设备 | |
CN111178277B (zh) | 一种视频流识别方法及装置 | |
CN110060264B (zh) | 神经网络训练方法、视频帧处理方法、装置及系统 | |
US11354544B2 (en) | Fingerprint image processing methods and apparatuses | |
WO2016192467A1 (zh) | 一种播放视频的方法及装置 | |
US20130051633A1 (en) | Image processing apparatus | |
KR102126370B1 (ko) | 동작 인식 장치 및 방법 | |
US9621780B2 (en) | Method and system of curve fitting for common focus measures | |
US9898828B2 (en) | Methods and systems for determining frames and photo composition within multiple frames | |
Ottenwälder et al. | Distributed complex event processing for mobile large-scale video applications | |
JP2017034616A (ja) | 画像処理装置及びその制御方法 | |
JP2010122840A (ja) | 物体領域検出方法、物体領域検出装置、物体領域検出プログラムおよびそのプログラムを記録したコンピュータ読み取り可能な記録媒体 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 13755128 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2014501999 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14379541 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 13755128 Country of ref document: EP Kind code of ref document: A1 |