A pedestrian detection method and system based on two-level attention mechanism

Abstract

The invention discloses a pedestrian detection method based on a two-level attention mechanism, which sends the original image into the RPN module to obtain target candidate frames; divides each candidate frame into three sub-regions from top to bottom; passes the sub-regions through the pool module The features of each candidate frame are sent to a first-level attention module to obtain relatively accurate part features in each sub-region; the part features of each candidate frame are stacked with the global features, and sent to a second-level attention module for global Adaptive weighting of features and part features; connecting two-level attention modules in series to obtain an end-to-end detection network model. The invention effectively removes complex background interference, is suitable for changing situations in real application environments, enhances detection robustness, and reduces false detection and missed detection probability, especially for blocked pedestrians and redundant noise interference with relatively little available information Serious small-scale pedestrians can effectively improve the detection ability of pedestrian targets in video images.

Description

technical field [0001] The invention relates to a method in the field of object detection in an image, in particular to a pedestrian detection method and system based on a two-level attention mechanism. Background technique [0002] The advent of the era of big data has promoted the continuous update and development of computer technology. Pedestrian detection technology, as a research hotspot in the field of computer vision, has shown important application value in the fields of intelligent video surveillance and intelligent transportation. Due to the following difficulties and challenges in the existing pedestrian detection algorithm, the detection results still need to be improved: due to the shooting distance, the picture is larger but the size of the target pedestrian is smaller, and the characteristics of the target area after the deep learning convolutional neural network is reduced Rarely, it is difficult to carry out effective detection and recognition; due to the f...

AI Technical Summary

Problems solved by technology

However, there are still some problems. On the one hand, this method still uses the features of the entire target candidate box, and does not fully extract the body part features. In addition, it pays too much attention to the body part features or the features of the entire candidate box, resulting in the network in the global The imbalance between local and local, so the detector does not have good generalization performance; on the other hand, using a part detector to extract body part features will introduce additional labeled body part information, resulting in an increase in cost

[0007] At present, there is no description or report of the similar technology of the present invention, and no similar data at home and abroad have been collected yet.

Images