Road scene image processing method based on double-side dynamic cross fusion

Abstract

The invention discloses a road scene image processing method based on double-side dynamic cross fusion. The method comprises two processes of a training stage and a testing stage, and includes selecting road scene images and corresponding thermodynamic diagrams and real semantic understanding images to form a training set; constructing a convolutional neural network; performing data enhancement on the training set to obtain an initial input image pair, and inputting the initial input image pair into a convolutional neural network for processing to obtain a corresponding road scene prediction map; calculating a loss function value between the road scene prediction map and the corresponding real semantic segmentation image; repeating the above steps to obtain a convolutional neural network classification training model; and inputting the road scene image to be subjected to semantic segmentation and the corresponding thermodynamic image into the convolutional neural network classification training model to obtain a corresponding predicted semantic segmentation image. According to the invention, the semantic segmentation accuracy of the road scene image is effectively improved, the loss of detail features is reduced, and the edge of an object can be better restored.

Description

technical field [0001] The present invention relates to a deep learning image processing method, in particular to a road scene image processing method based on bilateral dynamic cross fusion. Background technique [0002] The rise of the intelligent transportation industry has made semantic segmentation more and more widely used in intelligent transportation systems. From traffic scene understanding and multi-target obstacle detection to visual navigation, semantic segmentation technology can be used to achieve. Traditional semantic segmentation mainly relies on image texture, color, and other simple surface features and external structural features for image segmentation. The segmentation results obtained in this way are relatively crude, with low precision and no relevant annotations, that is, the image is only segmented into I have selected several blocks, but I don’t know the category of each block, so I need to specify it manually. The second is the semantic segmentati...

AI Technical Summary

Problems solved by technology

Almost all existing unmanned driving road scene understanding methods use deep learning methods, combining convolutional layers and pooling layers to build models. However, the features obtained by purely using pooling and convolution operations are not only single but also not representative. As a result, the extracted image feature information will be reduced, making the restored effect information rough and the segmentation accuracy low.

Images