Video snapshot method based on memory coding mechanism and video snapshot device thereof

Abstract

The invention discloses a video snapshot method based on a memory coding mechanism and a video snapshot device thereof. The method comprises the following steps that multiple frames of video images are continuously acquired according to a set snapshot period, and the timestamp of the snapshot moment is marked on the video images, wherein the video images include target characteristics; the target characteristics of the acquired multiple frames of video images are intelligently analyzed so as to select the metadata of the image frames having the optimal detection effect and the corresponding timestamp in a screening way; the acquired multiple frames of video images are coded, and the corresponding video images are searched from the coded video images according to the received timestamp to act as the optimal snapshot images; and the optimal snapshot images and the corresponding metadata are uploaded. According to the technical scheme, ineffective snapshot can be avoided, the images having the optimal snapshot effect can be acquired, the snapshot accuracy can be enhanced and the network bandwidth resource can also be saved.

Description

technical field [0001] The invention relates to the technical field of monitoring, in particular to a video capture method and device based on a memory coding mechanism. Background technique [0002] With the rapid development of the monitoring system, the scale of the system is getting bigger and bigger, and there are more and more monitoring points. The staff have no time to take care of more and more monitoring videos. Therefore, in order to reduce the cost of manpower and material resources, the intelligent front-end equipment Increasingly demanding. Capture cameras, as a solution, are also gaining more and more attention. The capture camera is a high-definition surveillance smart camera that can track, identify, and capture targets such as faces and vehicles. It not only covers the single video streaming function of traditional cameras, but also can directly detect targets such as faces and vehicles, and capture the required targets. Capture cameras can provide power...

AI Technical Summary

Problems solved by technology

[0003] At present, the camera system composed of several groups of capture cameras is commonly used, and can only capture targets such as faces and vehicles. However, how to avoid invalid capture and ensure that each capture is clear and accurate, and the metadata is accurate is the key issue in the capture system research.

Images