Image Preprocessing Algorithm Based on K-Means

Abstract

With the development of image acquisition and storage technology, the image data is greatly increased. How to process the increasing image data quickly has become the main problem of image processing. In this paper, the image data are processed by K-means algorithm in Python language environment, and the original image and the image processed by K-means algorithm are classified and trained in convolution neural network. The experimental results show that the time consumed by the image processed by K-means algorithm is 20 s to 30 s less than that of the original image in convolution neural network. It can effectively improve the efficiency of image processing.