Semantic segmentation method based on improved full convolutional neural network

Abstract

The invention discloses a semantic segmentation method based on an improved full convolutional neural network. The semantic segmentation method comprises the steps of: acquiring training image data; inputting the training image data into a porous full convolutional neural network, and obtaining a size-reduced feature map through a standard convolution pooling layer; extracting denser features while maintaining the feature map size through a porous convolutional layer; predicting the feature map pixel-by-pixel to obtain a segmentation result; using the stochastic gradient descent method SGD totrain parameters in the porous convolutional neural network in the training; acquiring image data that needs to be subjected to semantic segmentation, and inputting the image data the trained porous convolutional neural network, and obtaining a corresponding semantic segmentation result. The invention can improve the problem that the feature map of the final upsampling recovery in the full convolution network loses sensitivity to the details of the image, and effectively expands the receptive field of a filter without increasing the number of parameters and the amount of calculation.

Description

technical field [0001] The invention relates to a semantic segmentation method based on a porous fully convolutional neural network, which belongs to the field of computer vision. Background technique [0002] Image semantic segmentation is a key technology for image understanding, and it is widely used in automatic driving systems (specifically street view recognition and understanding), drone applications (landing point judgment) and wearable devices. Image semantic segmentation realizes the classification of all pixels in the image. Before deep learning is applied to image semantic segmentation, there are many methods such as the simplest pixel-level threshold method, segmentation method based on pixel clustering, and graph partitioning method. Shi et al. proposed a Normalized cut (N-cut) method based on graph partitioning, which takes into account the connection weights between different parts of the partition and the nodes of the whole graph to achieve the purpose of c...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to overcome the deficiencies of the prior art, provide a semantic segmentation method based on the improved full convolutional neural network, and solve the problem of continuous multiple maximum pooling and downsampling operations in the existing full convolutional network. The feature resolution is sharply reduced, and the feature map restored by the final upsampling loses the sensitivity to the details of the image.

Images