A Pedestrian Recognition Method Based on Human Head Features

Abstract

The invention provides a pedestrian recognition method based on human head features, which specifically includes the following steps: collect, define, learn and classify pedestrian contour maps with special features and low-frequency postures, and establish a posture template; use a laser scanner to obtain target objects The original image, and then preprocess the original image to obtain the contour map of the target object; use the ratio of the height and width of the circumscribed rectangle of the target overall contour map as the overall feature to classify the target; extract the target head contour according to the overall contour map Graph, its circularity is computed as a local feature for recognition to classify objects again. The present invention establishes a template for pedestrian postures with special features and low-frequency features, and utilizes the advantage that the head, shoulders and upper and lower parts of the legs are not easily blocked at the same time, avoiding the system being unable to identify pedestrians with special features; the present invention rounds the human head As the second-level classification of pedestrian recognition, the shape degree avoids the misidentification of the system and improves the pedestrian recognition rate.

Description

technical field [0001] The invention belongs to the field of perception technology, and in particular relates to a pedestrian recognition method based on human head features. Background technique [0002] In recent years, 3D laser scanning technology has developed rapidly. The 3D laser scanning system, also known as the 3D laser imaging system, is an advanced long-distance measurement technology, which can quickly and accurately obtain the spatial coordinate information of the object, and can store the collected data in the computer, and then quickly reconstruct The three-dimensional model of the object can be obtained to analyze and process it. The data information collected by the laser scanner is a point cloud composed of three-dimensional information and reflectivity of discrete data points. Each data point in the point cloud contains the distance value, angle value, reflectivity, etc. of the scanned point. Using this technology can realize the complete collection of ...

AI Technical Summary

Problems solved by technology

[0003] In pedestrian recognition, the recognition method of single-line laser radar is generally used. However, because single-line laser radar can only obtain a cross-section of the object, it cannot obtain more complete three-dimensional information.

Images