Research on Scene Semantic Segmentation Based on Deep Learning

Abstract

Because of the problem of the low accuracy and slow speed of the traditional semantic segmentation model, making it difficult to actually use. In response to this problem, this paper focuses on the method to improve the precision and speed of the algorithm. According to this theory, based on the convolution neural network, we have designed the PSPNet and ICNet models. Meanwhile, a scene semantic segmentation network based on deep learning was presented. The network effectively improves the accuracy of semantic segmentation of convolutional neural networks by merging multi-level depth and network features. The test results on the LISA traffic sign data set show that the proposed semantic segmentation network has outstanding performance compared with other state of the art semantic segmentation network structures.