Positive sample acquisition method, pedestrian detection model generation method and pedestrian detection method

Abstract

The invention discloses a method for obtaining a positive sample, a method for generating a pedestrian detection model, and a method for detecting a pedestrian. The method for obtaining a positive sample includes: adding disturbance to the original frame of the positive sample pre-marked in the original image to obtain a selection frame of the positive sample. The image surrounded by the original frame of the positive sample contains pedestrians; the image surrounded by the marquee frame of the positive sample is extracted as a positive sample. The technical solution of the present invention can effectively increase the number of positive samples available in the original image, and add a background to the positive samples, which can effectively improve the recognition accuracy of the trained pedestrian detection model.

Description

technical field [0001] The present invention relates to the technical field of intelligent identification, in particular to a positive sample acquisition method, a pedestrian detection model generation method, a pedestrian detection method, a storage device, and a pedestrian detection device. Background technique [0002] Pedestrian detection is an important research direction in the field of computer vision. By intelligently analyzing the visual images captured by the camera, pedestrians contained in the images are automatically detected. Pedestrian detection has a wide range of applications in the fields of vehicle assisted driving, intelligent monitoring, and intelligent robots. [0003] In the existing pedestrian detection process, the corresponding classifier is often trained based on pre-acquired positive samples and negative samples, and then the image to be detected is identified based on the classifier to detect whether the image to be detected contains pedestrians....

AI Technical Summary

Problems solved by technology

At this time, a pedestrian in the original image can only extract one positive sample, and the number of positive samples that can be extracted in the original image is limited, and it is difficult for the classifier trained based on the positive samples extracted by the positive sample identification box to contain the background Pedestrian images are recognized, resulting in low recognition accuracy

Images