A video concentration method and device

Abstract

The present application discloses a video concentration method and device, the method comprising: acquiring a video to be concentrated including multiple moving objects, and then selecting a combination method of moving objects for each concentrated image, each concentrated image obtained after concentrating the video to be concentrated Each frame image of each frame image, in order to carry out the video concentration of the video to be concentrated according to the combination mode of the moving target in each concentrated image. It can be seen that the present application can select a reasonable combination of moving objects for each frame of condensed image, so that each moving object in each frame of condensed image can occupy the image space to the greatest extent, and make the different moving objects in each frame of condensed image The degree of overlap between targets should be as small as possible, thereby improving the enrichment accuracy of the enriched video.

Description

technical field

[0001] The present application relates to the technical field of video processing, and in particular to a video concentration method and device. Background technique

[0002] With the improvement of people's security awareness, more and more attention has been paid to the safety of homes, campuses, transportation and public facilities, so a large number of surveillance videos have also been produced. Due to the limitation of storage size and scene complexity, there are great difficulties in the storage, transfer and screening of surveillance video. How to quickly and effectively obtain useful information from massive videos is a very important problem that needs to be solved. Therefore, video Enrichment technology came into being.

[0003] Video enrichment refers to a simple summary of the original video content, which is to rearrange the sequence of moving objects in time and space while retaining all moving objects (such as people and vehicles) in the orig...

Examples

no. 1 example

[0087] see figure 1 , which is a schematic flow chart of a video concentration method provided in this embodiment, the method includes the following steps:

[0088] S101: Acquire a video to be concentrated, where the video to be concentrated includes multiple moving objects.

[0089] In this embodiment, the original video that needs to be condensed is defined as the video to be condensed, such as surveillance video in some scenarios. The video to be concentrated is any video that may contain any moving objects, such as moving objects such as people and cars. Here, each moving object in the video to be concentrated is defined as a moving target.

[0090] In this embodiment, the background image in the video to be concentrated can be extracted through background modeling. Wherein, the background image refers to a clean background image after removing moving objects in the video to be concentrated, and the background image is obtained by continuously updating the background ima...

no. 2 example

[0151] It should be noted that this embodiment will introduce specific implementation manners of step S201 and step S202 in the first embodiment.

[0152] In this example, see image 3 In the schematic diagram of the clustering-based grouping process shown in the first embodiment, the step S201 of "obtaining each initial group by grouping overlapping moving objects in the video to be concentrated" may include the following steps S301-S302:

[0153] S301: Extract the clustering features corresponding to each moving target from the video to be concentrated.

[0154] In this embodiment, all overlapping moving objects in the video to be concentrated can be grouped by a clustering algorithm, so as to obtain each initial group, and each initial group includes one or more moving objects. In order to obtain each initial grouping through the clustering algorithm, it is necessary to extract the clustering features corresponding to each moving object from the video to be concentrated, t...

no. 3 example

[0187]This embodiment will introduce a video concentrating device, and for related content, please refer to the foregoing method embodiments.

[0188] see Figure 6 , which is a schematic composition diagram of a video concentrating device provided in this embodiment, the device 600 includes:

[0189] The video to be condensed acquiring unit 601 is configured to acquire the video to be condensed, the video to be condensed includes a plurality of moving objects;

[0190] The target combination selection unit 602 is configured to select a moving target combination mode for each condensed image, and each condensed image is each frame image obtained after condensing the video to be condensed;

[0191] The video condensing processing unit 603 is configured to perform video condensing on the video to be condensed according to the combination manner of the moving objects in each condensed image.

[0192] In an implementation of this embodiment, the target combination selection unit...

technical field

[0001] The present application relates to the technical field of video processing, and in particular to a video concentration method and device. Background technique

[0002] With the improvement of people's security awareness, more and more attention has been paid to the safety of homes, campuses, transportation and public facilities, so a large number of surveillance videos have also been produced. Due to the limitation of storage size and scene complexity, there are great difficulties in the storage, transfer and screening of surveillance video. How to quickly and effectively obtain useful information from massive videos is a very important problem that needs to be solved. Therefore, video Enrichment technology came into being.

[0003] Video enrichment refers to a simple summary of the original video content, which is to rearrange the sequence of moving objects in time and space while retaining all moving objects (such as people and vehicles) in the orig...

Examples

no. 1 example

[0087] see figure 1 , which is a schematic flow chart of a video concentration method provided in this embodiment, the method includes the following steps:

[0088] S101: Acquire a video to be concentrated, where the video to be concentrated includes multiple moving objects.

[0089] In this embodiment, the original video that needs to be condensed is defined as the video to be condensed, such as surveillance video in some scenarios. The video to be concentrated is any video that may contain any moving objects, such as moving objects such as people and cars. Here, each moving object in the video to be concentrated is defined as a moving target.

[0090] In this embodiment, the background image in the video to be concentrated can be extracted through background modeling. Wherein, the background image refers to a clean background image after removing moving objects in the video to be concentrated, and the background image is obtained by continuously updating the background ima...

no. 2 example

[0151] It should be noted that this embodiment will introduce specific implementation manners of step S201 and step S202 in the first embodiment.

[0152] In this example, see image 3 In the schematic diagram of the clustering-based grouping process shown in the first embodiment, the step S201 of "obtaining each initial group by grouping overlapping moving objects in the video to be concentrated" may include the following steps S301-S302:

[0153] S301: Extract the clustering features corresponding to each moving target from the video to be concentrated.

[0154] In this embodiment, all overlapping moving objects in the video to be concentrated can be grouped by a clustering algorithm, so as to obtain each initial group, and each initial group includes one or more moving objects. In order to obtain each initial grouping through the clustering algorithm, it is necessary to extract the clustering features corresponding to each moving object from the video to be concentrated, t...

no. 3 example

[0187]This embodiment will introduce a video concentrating device, and for related content, please refer to the foregoing method embodiments.

[0188] see Figure 6 , which is a schematic composition diagram of a video concentrating device provided in this embodiment, the device 600 includes:

[0189] The video to be condensed acquiring unit 601 is configured to acquire the video to be condensed, the video to be condensed includes a plurality of moving objects;

[0190] The target combination selection unit 602 is configured to select a moving target combination mode for each condensed image, and each condensed image is each frame image obtained after condensing the video to be condensed;

[0191] The video condensing processing unit 603 is configured to perform video condensing on the video to be condensed according to the combination manner of the moving objects in each condensed image.

[0192] In an implementation of this embodiment, the target combination selection unit...

Images