Image semantic segmentation method based on deep learning and storage medium

Abstract

The invention discloses an image semantic segmentation method based on deep learning and a storage medium. The image semantic segmentation method comprises the steps of connecting an average global pooling layer and a full connection layer in series after a feature extraction network to serve as a pre-training model for classification, and performing classification training on the pre-training model by adopting an Imagenet-1K data set; sequentially connecting the feature extraction network in the trained pre-training model with a lightweight ASPP module and two feature enhancement modules to form a semantic segmentation model; expanding the data set cityscapes through overturning, rotating and zooming, and training the semantic segmentation model by adopting the expanded data set to obtaina target semantic segmentation model; and inputting the preprocessed new picture into a target semantic segmentation model, carrying out forward propagation once in the target semantic segmentation model, and outputting a predicted semantic segmentation result in an end-to-end manner.

Description

technical field [0001] The present invention relates to image processing technology, in particular to an image semantic segmentation method based on deep learning and a storage medium. Background technique [0002] Most of the current best image semantic segmentation methods are mostly encoder-decoder frameworks based on deeplabv3+. Encoder part: First, deeplabv3+ obtains the feature extraction network by pre-training resnet on the ImagNet dataset, but downsampling will reduce the resolution of the feature, resulting in information loss, so the ordinary convolution of the last residual block is replaced by a hole convolution , each convolution within this residual block uses a different dilation rate to capture multi-scale contextual information. Then, the extracted features are input to the ASPP module. The ASPP module outputs the input features to five modules at the same time. The first module uses average pooling to fuse the features, so that the final feature map of t...

AI Technical Summary

Problems solved by technology

[0005] In view of the above-mentioned deficiencies in the prior art, the image semantic segmentation method and storage medium based on deep learning provided by the present invention solve the problem of large amount of computation in the image semantic segmentation method in the prior art

Images