Unlock instant, AI-driven research and patent intelligence for your innovation.

Image retrieval device, image retrieval method, and storage medium

An image detection and imaging technology, applied in the fields of digital data information retrieval, instruments, computing, etc., can solve the problems of searching for a large amount of data, such as the desired scene, and achieve the effect of improving the retrieval speed, reducing the retrieval noise, and reducing the operation.

Active Publication Date: 2019-04-19
HITACHI LTD
View PDF6 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, many existing security camera systems are systems composed of security cameras, video recorders, and players, and it is difficult to search for desired scenes from a large amount of accumulated data.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Image retrieval device, image retrieval method, and storage medium
  • Image retrieval device, image retrieval method, and storage medium
  • Image retrieval device, image retrieval method, and storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0035]

[0036] Figure 1A It is a configuration diagram of the video retrieval system 100 according to the first embodiment of the present invention.

[0037] The image retrieval system 100 is used to extract the object in the image specified by the user from a different time period of the image (for example, from a frame at a different time from the frame containing the object specified by the user), or from a different image (for example, from a frame at a time different from that containing the object specified by the user). The purpose of this system is to use the tracking information of the dynamic object (moving body) in the image to search for the query image suitable for retrieval. The query image is generated for each shooting location of the image to be searched, thereby improving the speed and accuracy of the search.

[0038] The video retrieval system 100 includes a video storage device 101 , an input device 102 , a display device 103 , and a video retrieval dev...

Embodiment 2

[0177] In the first embodiment, a method for improving the efficiency of similar image retrieval using tracking information of a moving object in a video is described. The tracking of dynamic objects is usually only based on the information between close frames. Therefore, when the object is stationary for a long time, or there is an obstacle between the object and the camera, the tracking is sometimes interrupted. If long-term tracking cannot be performed, the number of query candidates obtained from each moving line decreases, so that the effect of the first embodiment may not be fully exhibited.

[0178] Therefore, in the second embodiment, a method of correcting tracking information using salient regions stored in the image database 111 was described. Each part of the video retrieval system 100 of Embodiment 2 has the same characteristics as those shown in FIGS. Figure 12 Parts assigned the same symbols in the first embodiment shown have the same functions, and therefore...

Embodiment 3

[0216] In Embodiment 1, similar image retrieval is performed using salient regions of different frames on the same moving line, thereby improving retrieval accuracy. However, when there is no salient region characteristic of the object to be searched on the moving line, there may be limited images obtained by the search. In Embodiment 3, a method of notifying the user of a salient region of a search target object included in a video different from the video designated by the user will be described. Each part of the image search system 100 of Embodiment 3 has the same features as those shown in FIGS. Figure 12 Parts assigned the same symbols in the first embodiment shown have the same functions, and therefore descriptions thereof will be omitted.

[0217] Figure 17 It is an explanatory diagram related to the presentation of queries existing in different images according to the third embodiment of the present invention.

[0218] For example, in the explanatory diagram 1701 ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The video retrieval device of the present invention detects the movement of one or more moving objects from a first video consisting of a plurality of frames captured at a first point and a second video composed of a plurality of frames captured at a second point. path and stored in the storage device, extracting the image feature value of each frame of the selected moving body among the above-mentioned one or more moving objects detected from the first image and storing in the above-mentioned storage device, according to the above-mentioned The moving route of the selected moving object detected from the first image and the moving route of the one or more moving objects detected from the second image, and a query used as a search query are selected from the extracted image feature values. The image feature value retrieves the image feature value of the one or more moving objects extracted from the second image using the query image feature value, and outputs a result of the search.

Description

[0001] This application claims the priority of Japanese application Japanese Patent Application No. 2013-253897 for which it applied on December 9, 2013, The content is incorporated in this application by reference. technical field [0002] The invention relates to an image retrieval technology. Background technique [0003] With the popularization of security cameras, there is an increasing need to search for desired people, vehicles, etc. from images shot at multiple locations. However, many existing security camera systems are systems composed of security cameras, video recorders, and players, and it is difficult to search for a desired scene from a large amount of accumulated data. [0004] In this regard, attention is being paid to systems incorporating similar image retrieval technologies. If similar image retrieval techniques are used, it is possible to retrieve frames in which a specific person or object appears from a large amount of image information. Similar ima...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/00G06F16/583
CPCG06F16/583G06V20/40
Inventor 渡边裕树米司健一吉永智明广池敦
Owner HITACHI LTD