Video summary comparator
The video investigation system is designed for the public security front-line police to handle the case quickly at the scene. Through the system, the video investigators can quickly complete the whole series of process work, such as video collection, analysis, search, research and judgment, case management, etc. the operation is simple and practical, equipped with special protective box protection equipment, and the actual combat ability is excellent.
1. Core functions: Video concentration, video search, video enhancement, image clarity
2. Powerful functions: including case management, clue management, clue editing, report generation, concatenation and other functions
3. Excellent performance: leading industry in main technical performance indicators
The speed of video import is the fastest. 1g video is successfully imported in 5 seconds, which is the first video rectification technology;
Collision-free condensed video. Industry exclusive, ensure no target ghosting after video concentration;
The retrieval speed is the fastest. 50000 targets are retrieved in one second, leading other products in the industry.
Advantages of video detection system:
1. Close to actual combat: from video collection and management, case study and judgment, to clue labeling, case reporting, to meet the actual combat needs of police video investigation and case handling
2. Easy to use: easy to use video detection in 10 minutes
3. Fast experience: the AI technology engine with completely independent intellectual property rights can accelerate the operation through GPU, and the video image analysis and search speed is fast.
Video investigation workflow
1、 Two core steps of intelligent video detection:
Rough processing: find valuable clues through video concentration and fast video search technology in massive video;
Delicate processing: further enhance and clear the clues with poor video image quality;
2、 Main functions of video detection system
1. Case and event management, video investigation case database, clue database, case statistics, serial and parallel case functions;
2. Video concentration, video search, video nonlinear annotation, image clarity, video transcoding and other functions;
2. Video detection tools, universal play, electronic map, video platform download and call and other powerful auxiliary functions
Video image acquisition summary comparator
Video image acquisition summary comparator is a portable case handling equipment, which integrates three core functions of video acquisition, summary fusion and video retrieval as well as video investigation process. It solves the problems of slow video acquisition speed, heavy workload of video review, and complex search of video clues. At the same time, the video image acquisition summary comparator can use the network management module to provide the summary and retrieved video results to the client for review. It is suitable for 1 to 5 people to participate in the review of video, and can play the maximum performance of the video image summary comparator.
The network version of video image acquisition summary comparator is composed of mobile workstation (server) and call client, with the functions as follows:：
1. Video image acquisition summary comparator server: with case management, video summary compression processing, user management, summary review management and other functions;
2. Review client: it has video upload, review video, generate video clues, study and judge clues and other functions.
3. Mobile workstation (server) function:
Case management: it has the management of the name and description of the case and the monitoring point information associated with the case, which can realize the operation of adding, deleting and modifying.
Format compatibility: video summary processing function can compress one or more monitoring points, and can process at least 264、601、692、800、801、a64、aira、asf、av、avc、avi、avx、bn、bsr、csv5、dat、dav、djl、dvf、dvr、edh、exe、flv、ge5、h264、h3crd、h64、hav、he4、hie、ifv、ivs、lm、m4c、mg4、mis、mod、mp、mp4、mpg、mph、mpk、mts、nsf、nv4、Nvms_PlayHik、ps、rec、sdv、str、sv3、sv4、sv5、tk6、tkl、tm4、ts、vrd、vdg、ya2、zv and 80 kinds of video formats as well as 600 encoding formats, and the processed results are generated into index based on video content, supporting at least 4 video concurrent processing, and finally generating video files A brief summary video with standard encoding is provided to the client for review.
★ summary processing
1. Support the summary processing of1080P、720P、D1、CIF、2CIF、QCIF and other resolution videos;
2. Support video import while summary processing, and select whether to process the imported video immediately after import;
3. The video task of summary processing can be processed without sampling and twice sampling;
4. The task of summary processing can select two kinds of scenes for processing according to the video scene: day and night;
5. Support to switch and display case details, processing tasks and processing progress in the same interface in the form of switch tab; support to select cases in the list in the form of table to switch and display case details, processing tasks, processing progress and other information of the case;
★ electronic map management
1. It has plane and satellite maps of 19 levels and above, with map management function, which can be used by clients to browse maps, add, delete and modify mark point information, draw vector map trace, map survey, etc;
2. Support adding local pictures as case map;
3. Support manual addition of monitoring point marking information and base station information, and support batch import of at least 10000 monitoring point information at a time;
4. Support to draw track on offline map and save it on offline map;
5. Support the function of setting offline map as local map;
6. It supports the function of saving multiple track maps, exporting and deleting them in batch;
7. Support the automatic generation of tracks according to the lead time;
8. It supports selecting the rectangular area on the offline map through the mouse frame to control the display or hiding of monitoring points in the selected area on the offline map;
9. It supports dragging the monitoring point of the case into the offline map in the form of dragging from the case list on the left to mark the monitoring point;
Summary review management
With multi-user access, playback video at the same time, for the case involving more video, multi-user access can improve the efficiency of video access. In the process of video retrieval, various playback controls such as fast forward, slow playback, frame by frame, cycle, local zooming of the screen are provided, which can intercept clue screenshots and video clip.
★ density regulation parameters: density level: 1-5;
Multi screen playback window: ≥ 4;
Original target positioning speed index:
50,000-100,000 targets, speed ≤ 3 seconds;
10,000-50,000 targets, speed ≤ 2 seconds;
Less than 10,000 targets, speed ≤ 1s;
Note: the speed of original target location, that is, the length of time to retrieve the original video from the target in the summary file.
2. ★ concentrated play
The system shall have concentrated playback settings in the shortcut key below the video. After the area is delimited, the video will slow down the playback speed (0.2, 0.4, 0.6, 0.8, 1.0) for the target of the area of concern; for the target of the area of non concern, the video will play fast (1.0, 2.0, 4.0, 6.0, 8.0), and support multiple shortcut keys (frame by frame, fast forward, slow playback, screenshot, video clip download, Bureau Control of partial amplification, loop play, video adjustment, etc.)
3. Review task assignment
Through the distribution and management of multi-user rights, the review task is clear and clear, and the case video review task is standardized.
Video clue management
It has the function of summarizing trail information uploaded by each client, describing trail video, trail information and time calibration in case management, and displaying all trail information in the form of list. It can preview a single video clue, and use the time-space relationship for comprehensive analysis. The video clue list can be saved and printed.
★ video content retrieval management
1. Provide video retrieval according to the conditions of separation of people and vehicles, target color, intelligent rules, time domain retrieval, graphical retrieval, color separation, etc;
2. Support the export and secondary retrieval of retrieval results;
3. The original video screen supporting the retrieval results is a tracking frame displaying or hiding intelligent rules and marking targets;
4. Support the use of the same retrieval conditions to retrieve all video data of more than two cases, such as color, separation of people and vehicles, map line retrieval, etc;
5. The retrieval results can be sorted and displayed according to time, target size and similarity; at the same time, it supports the control of the number of images displayed in each row, at least 4, 6, 8 and 10 images per row;
6. Support to display search results by entering page number jump;
Case statistics managemen
a) Case information statistics: Statistics of case videos, clues, review and other quantitative information, to facilitate the filing of case information and regular case data analysis.
b) Case summary report: summarize the whole clue information of the case, automatically generate the document of case report, so as to facilitate the summary of case settlement information by the case investigators and the sharing of case information in the case progress stage.
c) ★ one click summary to the system's own case video database, including the original video, summary video and clue video. The database field description requirements are formulated in accordance with the public security industry standard GA / Z 1129-2013.
d) ★ support the automatic serial parallel comparison function of cases, query according to various conditions, and provide various means to assist in the research and judgment of intelligence information.
It has the function of setting the cases displayed in case management and controlling whether they can be deleted; the checked cases cannot be deleted in case management;
Case setting and basic setting
1. It supports parameter setting to control the number of CPU threads that can be used by the device;
2. It can control the minimum remaining space of the summary file storage location;
3. It can control the total file size of each monitoring point task;
4. You can control which cases are displayed in the case management interface;
5. It can control the number of seconds of picture playing in advance and the total number of seconds of picture playing;
★ two / three dimensional model comparison function
Provide the function of vehicle model comparison between two / three-dimensional and video screenshots. The two / three-dimensional model can be used to build a database, with the number of two-dimensional model pictures ≥ 8000, and the number of three-dimensional model models ≥ 450. The model shall be continuously updated and upgraded free of charge.
For the videos that cannot be summarized, that is, some videos with unrecognized format, the client provides the function of copying the screen, recording the videos with unrecognized format on the screen, and converting them into the general encoding format that can be recognized by the system. This function is convenient for the investigators to convert the video format when they rest at night.
Clue video editing
Users can edit the locally cut cue video, the so-called non-linear editing. After each clip of cue video is edited separately, it is synthesized according to the sequence of the cues, and spliced into a completed dynamic track video. At the same time, the video can be exported to local and secondary editing.
The clues found by reviewing the summary video locally can be summarized in the online query generation function, and the functions such as video preview, thread description information editing, video export, manual addition of clues, and export of clue information list can be performed.
User management: it has the management function of the client user, and defines different permissions according to the user role.
Video and picture processing: it can clear the screenshots and dynamic video clips of the video, including sharpening, image enhancement, contrast adjustment, single frame reconstruction, histogram adjustment and other functions.
Online upgrade function: in the public security network or the Internet, the software program can be upgraded automatically.
★ introduction to use: it supports the use help in the form of video in the main interface, at least with the help of video demonstration forms such as how to establish a case, how to import video, how to process video, how to view video summary, how to conduct intelligent rule retrieval, how to generate clues, how to use maps, etc;
Accessing client functions
Abstract access: each client has access to the summary video processed on the mobile workstation (server) function, and carries out picture annotation, video cutting, clue information description, etc. for the discovered target. The client can mark the access progress of the summary video, so as to facilitate the task details during the handover.
★ offline electronic map: the client has the function of browsing electronic map. The offline electronic map service comes from the server (i.e. video image acquisition summary comparator), including monitoring point marking, map measurement, track drawing, base station marking, other geographic information point marking, two-dimensional map display and other functions of offline map.
Video upload: each client has the function of uploading video. It uploads the video resources submitted by the collector, stores them on the video cloud backup device (storage server) and manages them centrally. It also has the video image acquisition summary comparator to process the video centrally. At the same time, other clients can retrieve the existing original video resources from the backup device at any time.
★ target screening: each client can carry out target screening locally, that is, retrieval processing. The optional conditions include target color, active area, cross-line detection, and the realization of separate retrieval of people and vehicles.
Picture playing function: each client has the picture playing function, which can extract the key frames in the original video, display the pictures in the way of one second and one frame, and double-click the pictures to play the corresponding original video.
Case information upload: the device can upload the case information in the system to the video investigation online combat system or the involved video database, so as to realize the sharing and release of case information with the system.
License plate recognition and retrieval: the device has the function of automatically intercepting license plate information from video and displaying all license plate information in the form of pictures. It can perform accurate and fuzzy retrieval of license plate number. The retrieved results can be associated with the original video for playback and can generate video clues.
Face recognition retrieval: the device can automatically extract all the face information in the video, and show it in the form of pictures. It can input an external face image, and retrieve the extracted face image. The retrieval results can be sorted by similarity.
Clue screening sharing: the clue screening function refers to that the client user can preliminarily screen the local found clue information, mark the urgency of the clue, facilitate the secondary analysis and judgment of the upper authority, and preview, play and export the information of the clue during the screening process.
Two / three-dimensional vehicle model comparison function: the equipment can provide vehicle model comparison function between two / three-dimensional and video screenshots. The vehicle model can be built with two / three-dimensional model database. The number of two-dimensional vehicle model pictures ≥ 10000, and the number of three-dimensional vehicle model ≥ 50, and the vehicle model shall be continuously updated and upgraded free of charge.
Original video reading: the client can play the original video and provide the function of playing the original video in four windows at the same time. In the process of playing, the client can also take screenshots, annotate, cut the video and describe the text clues of the objects in the video.
No. 8 Lianyou Road, Chengnan District, Jingjiang, Jiangsu, China