An image processing method based on a neural network system and the neural network system

Abstract

The invention provides an image processing method based on a neural network system and the neural network system, and the method comprises the steps: enabling the neural network system to identify a current input image frame, and judging whether the current input image frame is a basic image frame or not; if the basic image frame is the basic image frame, performing calculation processing by the multi-layer convolutional layer of the neural network; if not, determining that the image is a basic image frame; if so, comparing the image data with the image data of the basic image frame to extractimage data different from the basic image frame; and performing calculation processing on the basis of the image data by the plurality of convolution layers, and finally, replacing the image data ofthe corresponding position of the intermediate complete data stored in the layer with the data calculated and output by the last convolution layer of the neural network by the last convolution layer of the neural network to obtain the image data of the new complete image. According to the method provided by the invention, when the neural network system processes data, unchanged information in front and back frames can be hidden, and only the changed information is processed, so that a large amount of computing resources can be saved, and the processing speed is increased.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to an image processing method based on a neural network system and a neural network system. Background technique [0002] In today's era, with the surging development of artificial intelligence technology, it not only affects people's production and life in all aspects, but also promotes the development and progress of the world. In order to effectively utilize the information collected from each terminal, the requirements for real-time processing of information are also constantly increasing. Especially in the field of security, the input of video information is large, and the amount of calculation is large, resulting in high power consumption. However, in most cases, there is not much information difference between the front and back frames of the video information. For the traditional solution, when processing video frames, regardless of whether there is any change in t...

AI Technical Summary

Problems solved by technology

Especially in the field of security, the input of video information is large, and the amount of calculation is large, resulting in high power consumption

However, in most cases, there is not much information difference between the front and back frames of the video information. For the traditional solution, when processing video frames, regardless of whether there is any change in the information between frames, all the data of each frame will be input into the neural network. System and other algorithms to process, resulting in the need to process a lot of repeated data

Images