Street scene analysis semantic segmentation method for automatic driving

Abstract

A street scene analysis semantic segmentation method for automatic driving comprises the following steps that an image semantic segmentation network is constructed, the image semantic segmentation network is used for performing down-sampling processing on an image to obtain an initial feature map, and one-dimensional horizontal pooling processing is performed on the initial feature map to obtain a first global feature map; performing one-dimensional vertical pooling processing on the initial feature map to obtain a second global feature map, fusing the first global feature map and the second global feature map, and generating an output image; collecting a training picture, and training the image semantic segmentation network by using the training picture; and performing semantic segmentation on a to-be-processed image by using the trained image semantic segmentation network. The long and narrow pooling mode of a global one-dimensional pooling mechanism is utilized, all information in the horizontal direction and the vertical direction can be directly and effectively aggregated, a large amount of information is associated, effective context information is formed, and the defects of traditional rectangular pooling in long-distance context information aggregation are overcome.

Description

technical field [0001] The invention relates to the field of image semantic segmentation, in particular to a street scene analysis semantic segmentation method for automatic driving. Background technique [0002] Due to the continuous increase of car users, road traffic congestion, safety accidents and other problems are becoming more and more serious. With the support of Internet of Vehicles technology and artificial intelligence technology, autonomous driving technology can coordinate travel routes and planning time, thereby greatly improving travel efficiency and reducing energy consumption to a certain extent. For fast visual tasks such as autonomous driving, both the accuracy and efficiency of image semantic segmentation are very important, but the current semantic segmentation network cannot achieve a good balance between the two. [0003] At present, in order to improve the efficiency of semantic segmentation, a large number of lightweight network researches applied ...

AI Technical Summary

Problems solved by technology

[0004] However, the existing network uses a two-dimensional square pooling operator to aggregate local area features to achieve feature pyramids and pyramid pooling modes with consistent size and proportion. This square pooling mode with a pooling aggregation range of a square area can only Aggregating object information in a local area cannot obtain effective long-distance context information

In addition, for some objects with irregular shapes and sizes, such as trees and utility poles, the two-dimensional square pooling operator will inevitably introduce irrelevant noise information, which will affect the network's overall understanding and judgment of features.

Images