A video image processing method suitable for video surveillance equipment

Abstract

The invention discloses a video image processing method suitable for video monitoring equipment. The method is used to obtain an image frame sequence; split the image frame sequence into a training set and a test set, and train a neural network model through the training set to obtain a trained neural network model. Neural network; extract the foreground image Extract the foreground image of each image frame to generate the fuzzy sequence of the foreground image of each image frame in turn, and judge whether each image frame has a fuzzy area in turn according to the corresponding fuzzy sequence, if yes, delete the image frame with the fuzzy area, Otherwise, saving image frames without blurred areas can compress the volume of video or image sequences composed of multiple images, so that on video surveillance equipment, the storage space required by the server can be greatly compressed, and the proportion of valid data can be increased. Clean up invalid and fuzzy data to improve the user experience.

Description

technical field [0001] The invention belongs to the technical field of image processing and mobile video processing, and in particular relates to a video image processing method suitable for video monitoring equipment. Background technique [0002] When video surveillance equipment collects videos of high-speed moving objects, most of the images obtained are very blurry, or images in dimly lit environments, or exposed images under strong light, and in long-term continuous video data After the collection, a large amount of these invalid data is accumulated, and the storage space consumed by the video surveillance equipment on the server side is huge. If these relatively blurry images can be removed, the volume of the video or the image sequence composed of multiple images can be compressed. , so that on the video surveillance equipment, the storage space required by the server can be greatly compressed, the proportion of valid data can be increased, the invalid and fuzzy data...

AI Technical Summary

Problems solved by technology

[0002] When video surveillance equipment collects videos of high-speed moving targets, most of the acquired images are very blurred, or are images in dimly lit environments, or exposed images under strong light. In the long-term continuous video data After collection, a large amount of these invalid data have been accumulated, and the storage space consumed by the video surveillance equipment on the server side is massive. If these relatively blurred images can be removed, the volume of the video or the image sequence composed of multiple images can be compressed , so that on the video surveillance equipment, the storage space required by the server can be greatly compressed, the proportion of valid data can be increased, invalid fuzzy data can be cleaned up, and the user experience can be improved. However, existing image processing methods do not There is no similar technology to solve the above problems. It is very time-consuming and labor-intensive to manually clean up unnecessary images or edit unrecognizable images of video frames.

Images