Pedestrian detection method fusing depth perception features and kernel extreme learning machine

Abstract

The invention discloses a pedestrian detection method fusing depth perception features and a kernel extreme learning machine. The method comprises the following steps: 1, constructing a DAGnet neuralnetwork comprising two parts; 2, training a DAGnet convolutional neural network by using the preprocessed sample to obtain a DAGnet model; 3, obtaining a depth perception feature vector by using a DAGnet model; 4, training the kernel extreme learning machine by using the depth perception feature vector to obtain a pedestrian recognition model; 5, performing generalization performance estimation onthe kernel extreme learning machine; 6, learning a second-level feature map and a GVBS saliency detection algorithm by using a DAGnet model to obtain a saliency map of the test image, and marking anapproximate area of a pedestrian in the test image; 7, scanning the approximate region by using a multi-scale sliding window to obtain a depth feature vector of the region where the window is located;and 8, identifying whether the area contains pedestrians or not by utilizing a pedestrian identification model. The method can obtain better detection performance, effectively improves the precisionand speed of pedestrian detection, and is better in robustness.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a pedestrian detection method integrating depth perception features and kernel extreme learning machines. Background technique [0002] With the development of unmanned driving, pedestrian detection, one of the important technologies in the intelligent traffic recognition system, has great research value. The purpose of the pedestrian detection algorithm is to detect pedestrians appearing in the image in real time, and use a rectangular frame to give the pedestrian's position. However, in different scenes, the illumination, diversity of pedestrian postures, shooting angles, and other objects similar to pedestrians in the scene make pedestrian detection technology face huge challenges. [0003] Pedestrian detection technology is mainly divided into two modules: target feature extraction and classifier design. The features applied to pedestrian detection mainl...

AI Technical Summary

Problems solved by technology

However, as the network progresses layer by layer, some significant features will be lost, and the classifier composed of fully connected layers in the convolutional neural network algorithm has weak generalization ability and slow calculation speed.

In the traffic monitoring system, the accuracy and detection speed are the key to evaluate the success of an algorithm. Many current algorithms stay in the network structure to deepen the length or thicken horizontally, resulting in cumbersome network structure, complicated training, and slow test speed.

Images