Cityscape image semantic segmentation method based on multi-feature fusion and Boosting decision forest

Abstract

A cityscape image semantic segmentation method based on multi-feature fusion and Boosting decision forest includes the following steps of carrying out super-pixel segmentation on images, carrying out multi-feature extraction, carrying out feature fusion and carrying out training learning and classification recognition. The method effectively integrates 2D features and 3D features and remarkably improves recognition rates of targets. Compared with the prior art, segmentation results are consistent, connectivity is good, edge positioning is accurate, a Boosting decision forest classification mechanism is introduced, and stability of target classification is guaranteed.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to a method for semantic segmentation of street view images based on multi-feature fusion and Boosting decision forest. Background technique [0002] Image segmentation is the technology and process of dividing an image into multiple regions with similar characteristics, and it is an important issue in image processing. The features here can be grayscale, color, texture, etc. of pixels, and the predefined target can be a single area or multiple areas. Image segmentation is not only the basis of object representation, but also has an important impact on feature quality, and can transform the original image into a more abstract form, making higher-level image analysis and understanding possible. Image understanding in computer vision, such as object detection, object feature extraction, and object recognition, etc., all rely on the quality of image segmentation. Image proce...

AI Technical Summary

Problems solved by technology

Since each pixel needs to be judged, the pixel-level target recognition algorithm will also generate a huge amount of data calculations. At the same time, most of the neighborhoods of adjacent pixels overlap, so the extracted features are also relatively similar. The final category judgment The results are not much different, but a large amount of redundant data is included in the calculation process, which greatly increases the amount of calculation

Images