Grayscale image colorization method based on convolutional neural network

Abstract

The invention discloses a gray image colorization method based on a convolutional neural network. The gray image colorization method comprises the steps: building the convolutional neural network, andconverting a gray image into a color image, wherein the hidden layer of the convolutional neural network comprises a plurality of connecting layers, and each connecting layer comprises a convolutional layer, a batch standardization layer and a combined nonlinear activation function layer which are connected in sequence; and the combined nonlinear activation function layer comprises a nonlinear activation function layer, a single-channel convolution kernel layer, a batch standardization layer and a normalization layer which are connected in sequence; and the combined nonlinear activation function layer performs nonlinear activation processing on a result after convolution operation in a feature layer-by-feature layer manner. The gray image colorization method has the advantages of automatic colorization, large application scene and the like, and can realize the function by using a small number of layers, and the colorization effect is better than that of the traditional method.

Description

technical field [0001] The invention relates to a grayscale image colorization method, in particular to a grayscale image colorization method based on a convolutional neural network. Background technique [0002] Currently, grayscale images are single-channel images with only one gray level. There are various applications of grayscale images in social life. In addition to the most common visible light that can be formed into grayscale images, images formed by other non-visible light including gamma rays, X-rays, ultraviolet light, infrared light, and microwaves are all There is only one single-channel image with light intensity, and these grayscale images are widely used in various fields such as medical diagnosis, industrial inspection, astronomical observation, and military investigation. Although grayscale images are widely used, compared with common color images, grayscale images have lower resolution, lower contrast, blurred edges, poor visual effect, difficult to iden...

AI Technical Summary

Problems solved by technology

Although these methods based on convolutional neural networks can achieve the goal of colorizing grayscale images, in the process of extracting grayscale images, these methods perform upsampling in order to restore the size of the image, making The image will lose local information, and the number of layers of these methods is relatively small, which may easily cause problems such as insufficient extraction of image features.

[0011] From the above analysis, it can be seen that the colorization of grayscale images is an important aspect of image processing. The traditional methods in the past are difficult to meet the requirements. With the development of convolutional neural networks, domestic and foreign scholars have made some attempts and made some progress. , but there is still room for improvement in the convolutional neural network model used for colorization and the results of processing

Images