Image semantic segmentation method, device, computer equipment and storage medium

Abstract

The present application relates to an image semantic segmentation method, apparatus, computer equipment and storage medium. The method includes: discretizing the depth image corresponding to the image to be processed according to a preset scene resolution, and determining the pixel regions of the prediction branches at all levels; determining the context of the convolution feature map corresponding to the image to be processed in the prediction branches at all levels information; according to the context information of the convolution feature map in the current prediction branch and the enhanced feature map in the superior prediction branch, obtain the enhanced feature map of the convolution feature map in the current prediction branch; according to the enhanced feature map and pixel area of ​​the prediction branch at all levels Perform classification prediction to obtain a segmented image of the image to be processed. Through this cascading method, the context information of the upper-level prediction branch can be transferred to the enhanced feature map of the lower-level prediction branch, which enriches the context information of the enhanced feature maps in the prediction branches at all levels, thereby improving the utilization of the enhanced feature maps at all levels. The accuracy of making classification predictions.

Description

technical field [0001] The present application relates to the technical field of image segmentation technology, and in particular, to an image segmentation technology method, apparatus, computer equipment and storage medium. Background technique [0002] Image semantic segmentation is an important research topic in the field of computer vision and pattern recognition, and is widely used in AI (Artificial Intelligence, artificial intelligence) scenarios such as autonomous driving systems and drones. It is divided into regional blocks with certain semantic meaning, and the category of each regional block is identified, and finally a segmented image with semantic annotation is obtained. [0003] In the existing research on image semantic segmentation, attempts are made to use the depth information of images to assist semantic segmentation. However, there is almost no correlation between depth information and color channels used in semantic segmentation, and the existing segmen...

AI Technical Summary

Problems solved by technology

However, there is almost no correlation between the depth information and the color channels used in semantic segmentation, and the existing segmentation effect of using depth information to assist semantic segmentation still needs to be improved

Images