Infrared road scene segmentation method based on category prototype regression

Abstract

The invention relates to an infrared road scene segmentation method based on category prototype regression, and the method comprises the following steps: 1, category prototype feature regression: obtaining a category feature prototype through regression through a large number of data tags and depth features, 2, construction of a relation matrix: after obtaining a category feature prototype, constructing the relation matrix through depth features and the category feature prototype, 3, attention enhancement: constructing different attention graphs through the relation matrix to realize feature enhancement, and 4, construction of attention modules: establishing a category attention module and a space attention module, and aggregating the functions of the two attention modules. According to the method, a strategy of category prototype regression is proposed to perform regression on the whole data set, representative category prototype features are obtained, and meanwhile, network depth features are clustered, so that global category features are tighter, and meanwhile, the difference between the categories is amplified, and a relation matrix and an attention module are correspondinglyconstructed, so that the overall characteristics are tighter, and the final image segmentation precision is improved.

Description

technical field [0001] The invention relates to an infrared road scene segmentation method based on category prototype regression, and belongs to the technical field of image processing. Background technique [0002] Compared with other scenes, vehicle road scenes are more complex, and many problems may occur in complex scenes, such as complex backgrounds that make target recognition more difficult, or similarities between targets that interfere with visual features , Different targets are misclassified, especially in the infrared vehicle road scene, the edge of the target is weak, and the boundary line between the background and the foreground is not obvious, which will lead to the accuracy of visual features. Therefore, in order to achieve higher recognition accuracy, the segmentation model needs to have stronger discriminative ability for weak edges and similar objects. At present, image semantic segmentation technology is mainly aimed at classification tasks at the pixe...

AI Technical Summary

Problems solved by technology

[0004] At present, the most complex vehicle-mounted road scene is the urban street scene, because in the street, the scene is complex and changeable, the background and the target are mixed together, it is difficult to distinguish, and the traffic volume on the road is large, sometimes there is dense situation, the general algorithm is difficult To achieve higher precision, at present, it is difficult for intelligent driving to achieve zero errors

Images