A dilated fully convolutional neural network device and its construction method

Abstract

The invention discloses an expanded fully convolutional neural network and a construction method thereof. The neural network includes a sequentially connected convolutional neural network, a feature extraction module, and a feature fusion module. The construction method is as follows: Select the convolutional neural network: remove the fully connected layer and classification layer used for classification in the convolutional neural network, leaving only the middle convolutional layer and pooling layer, and remove the convolutional layer and pooling layer Extract the feature map; construct the feature extraction module: the feature extraction module includes a plurality of series-connected expansion upsampling modules, and each expansion upsampling module includes a feature map merging layer, an expansion convolution layer and a deconvolution layer; construct Feature fusion module: The feature fusion module consists of a dense dilated fusion convolution block and a deconvolution layer. The invention effectively solves the problem of feature extraction and fusion in the convolutional neural network, and can be applied to the pixel-level labeling task of images.

Description

technical field [0001] The invention belongs to the technical field of image signal processing, in particular to an expanded full convolutional neural network device and a construction method thereof. Background technique [0002] Convolutional Neural Networks (CNNs) are the most widely used deep learning networks in image processing and computer vision. CNN was originally designed for image recognition and classification, that is, after the input image is passed through CNN, the category label in the output image is output. However, in some areas of image processing, it is not enough to just identify the category of the entire image. For example, in image semantic segmentation, it is necessary to label the category of each pixel in the image. At this time, the output is not a category label, but a map with the same size as the original image. Each pixel in the map is The semantic category to which the corresponding pixel in the original image belongs is marked. At this t...

AI Technical Summary

Problems solved by technology

[0005] However, the result map obtained from the previous fully convolutional network often cannot preserve the edge information of the object well, and the result map is often rough. Generally, a post-processing process is used to improve the labeling accuracy.

These post-processing processes not only increase the complexity of the labeling model, but because the labeling process is artificially divided, the results obtained are not smooth, and there are many discontinuous pixels, which have a great impact on the results.

These shortcomings are mainly because the previous FCN did not extract and utilize the image features in the network well, resulting in a decline in the resulting performance

On the other hand, the previous FCN has a large amount of parameters, which is not conducive to the transplantation and miniaturization of the model.

Images