Convolutional neural network road scene classification and road segmentation method

Abstract

The invention relates to a convolutional neural network road scene classification and road segmentation method. A dual-task joint structure model based on the convolutional neural network is established.and composed of an encoder and two decoders. Feature information sharing is achieved through end-to-end training.Classification of various road scenes such as urban roads, rural roads and highwaysis completed, road drivable areas in the scenes are segmented, and the dual goals of road scene classification and road area segmentation are achieved. Based on a convolutional neural network, scene classification and road extraction in road scene perception are combined. Real-time output is obtained through an end-to-end training mode, the functions of an intelligent driving assistance system areimproved, and an effective and reliable driving assistance function is provided. A mode of combining a full convolutional network and a conditional random field is adopted for a part extracted from adrivable area in the model, a high-resolution image optimization output result can be generated, and a relatively accurate segmentation effect is achieved.

Description

technical field [0001] The invention relates to an image recognition technology, in particular to a convolutional neural network road scene classification and road segmentation method. Background technique [0002] The perception of the road environment is a key technology in vehicle assisted driving technology, and the analysis and understanding of road scenes is an important part of the vehicle intelligent system. In terms of autonomous driving, the goal of road scene recognition is to be able to automatically obtain high-level semantic information of road images, judge the scene category to which the image belongs, and extract the drivable area in the image. For example, in the context of complex conditions such as rain, snow, and smog, classifying and identifying road scenes such as streets, country roads, and expressways, and giving early warning prompts such as speed adjustments to driving vehicles will help realize unmanned driving. In terms of automotive assisted dr...

AI Technical Summary

Problems solved by technology

[0008] The present invention is aimed at the problem of multi-type complex road scene recognition, and proposes a convolutional neural network road scene classification and road segmentation method. In the method, a dual-task joint structure model based on convolutional neural network is established. An encoder, Composed of two decoders, through end-to-end training, feature information sharing is realized, and various road scenes such as urban roads, country roads and expressways are classified, and the drivable area of ​​the road in the scene is segmented to achieve road scene Dual Objectives of Classification and Road Region Segmentation

Images