A pedestrian detection method based on deep learning multi-network soft fusion

Abstract

The invention discloses a pedestrian detection method based on deep learning multi-network soft fusion, and relates to the technical fields of image processing, target detection and deep learning; it includes S1: inputting an image to be processed; S2: inputting the image to be processed into a Darknet-based In the YOLO v3 pedestrian candidate area generator with 53 as the basic network, pedestrian candidate areas are generated; S3: input the image to be processed into the front-end prediction module, and output C feature maps; S4: input C feature maps into the semantic segmentation system, and output C A feature map containing context information; S5: Fuse the results of the semantic segmentation system with the pedestrian candidate results generated by the pedestrian candidate area generator; S6: Output the detection image. The present invention parallel softly fuses two systems of pedestrian candidate area generator and semantic segmentation, efficiently detects pedestrians in various challenging scenarios, and improves the detection ability of small targets at the same time.

Description

technical field [0001] The invention relates to the technical fields of image processing, target detection and deep learning, in particular to a pedestrian detection method based on deep learning and multi-network soft fusion. Background technique [0002] Object detection is an important problem in computer vision, which requires detecting the location of objects in video or digital images. Object detection is widely used in image detection, object recognition, video surveillance and other fields. Pedestrian detection, as a branch of object detection problems, involves detecting specific human categories, and it has a wide range of applications in areas such as autonomous driving, person recognition, and robotics. [0003] The goal of pedestrian detection algorithms is to draw bounding boxes in images or videos that accurately describe the location of pedestrians in real time. However, this is difficult to achieve due to the trade-off between accuracy and speed. Because ...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is: the present invention provides a pedestrian detection method based on deep learning multi-network soft fusion, which overcomes the fact that the existing method cannot accurately describe the pedestrian position in real time under the trade-off between pedestrian detection accuracy and speed The problem can improve the detection ability in the case of real-time detection