Context aggregation residual single image rain removal method based on convolutional neural network

A convolutional neural network and a single image technology, applied in the field of image processing, can solve problems that affect the performance of computer vision systems, image visual effects and image quality, and achieve good image rain removal effects, rich details, and simple implementation Effect

CN112184573AActive Publication Date: 2021-01-05XIAN UNIV OF TECH
6 Cites 7 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Publication Date
2021-01-05

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

A context aggregation residual single image rain removal method based on a convolutional neural network comprises the following specific steps: step 1, preprocessing an image; 2, constructing a convolutional neural network model; 3, training the convolutional neural network model obtained in the step 2 by adopting the preprocessed image obtained in the step 1 to obtain a rain removal convolutionalneural network model, constraining the rain removal convolutional neural network model by utilizing a loss function, and then performing back propagation to update parameters to obtain a trained rainremoval network model; and 4, inputting an image to be processed with rain into the trained rain removal network model, and finally outputting the rain-removed image. Due to the fact that the convolutional neural network model is constructed, more detailed features are obtained, more details are obtained, implementation is easy, and the image rain removing effect is good.
Need to check novelty before this filing date? Find Prior Art

Description

technical field

[0001] The invention relates to the technical field of image processing, and relates to a context aggregation residual single image rain removal method based on a convolutional neural network. Background technique

[0002] With the rapid development of science and technology, human society is entering the information society, the application of computers is becoming more and more extensive, and image processing technology is becoming more and more important to the development of various fields. According to statistics, when humans obtain external information, more than 70% of them come from vision, and image information has become the main means for humans to obtain information. With the progress of society and the development of economy, human vision is far from being able to meet the needs of information processing in some areas of daily life. Especially in recent years, with the image processing technology and computer vision technology [1] The rapid dev...

Examples

Embodiment 1

[0092] Such as figure 1 and figure 2As shown, the present invention is based on the convolutional neural network context aggregation residual single image deraining method comprising the following steps:

[0093] Step 1: Preprocess the input image, normalize the pixel value of the image to [0,1], and crop it to 256x256x3.

[0094] Step 2: Construct a convolutional neural network model;

[0095] The principle of constructing a convolutional layer: use convolution operation, instance normalization and activation function ReLU to combine into a convolutional layer,

[0096] F=ReLU(Instance_norm(Conv(x))) (5);

[0097] Context aggregation module DCA_Block: In the writing of the experimental code, the DCA_Block module is encapsulated into a function, so that this function can be called directly when this module is needed when the network is written later. The context aggregation module uses convolution with different expansion rates to obtain different feature maps; the contex...