Object Detection Method Based on Super Feature Fusion and Multi-Scale Pyramid Network

Abstract

The invention relates to a target detection method based on super-feature fusion and multi-scale pyramid network, which includes using deep convolutional neural network to extract layered multi-scale feature maps with different feature information; performing super-feature fusion; constructing a new multi-scale pyramid Network; construct target candidate boxes of different sizes and aspect ratios according to different layers; construct a new convolution module for multi-feature extraction and prevent gradient disappearance; use multi-task loss function for multi-category classifiers and bounding boxes The regressors are jointly trained and optimized to achieve image classification and target location functions. The present invention utilizes the feature extraction ability of the deep convolutional network for the target, considers the super-feature fusion method to improve the feature expression ability, generates a new module to prevent the gradient from disappearing and can help training and feature extraction more effectively, and constructs a target detection system The fully convolutional neural network improves the detection accuracy of the algorithm and obtains good target detection results.

Description

technical field [0001] The invention belongs to the technical field of computer vision target detection, in particular to a target detection method based on super-feature fusion and multi-scale pyramid network. Background technique [0002] The purpose of computer vision research is to use computers to realize human perception, recognition and understanding of the objective world. Object Detection is the most common technology in computer vision, and has received extensive attention in the field of computer vision theory research, and has broad application prospects. As one of the core research topics in the field of computer vision, target detection technology extracts target features through analysis, and then obtains the category and location information of the target. Object detection technology integrates cutting-edge technologies in many fields such as image processing, pattern recognition, artificial intelligence, computer vision, etc. Wide range of applications. ...

AI Technical Summary

Problems solved by technology

In 2016, the University of Washington team proposed a new target detection method YOLO, which solved the entire target detection process as a regression problem. YOLO detection speed is fast, but the accuracy is lower than the method based on region candidates.

[0006] In summary, although the target detection algorithm has achieved good results after decades of development, and the emergence of convolutional neural networks has improved the target detection accuracy a lot, many problems still need to be improved, for example, how to improve Effectively enrich target feature information, how to fuse features and solve the problem of gradient disappearance that may occur in deep convolutional neural network training, etc.

Images