Pedestrian detection method based on deep convolutional neural network

Abstract

The invention discloses a pedestrian detection method based on a deep convolutional neural network. The pedestrian detection method comprises the following steps of 10 sample set collecting, wherein sample set images are divided into training set images and test set images; 20 sample image preprocessing, wherein the sample set images are subjected to size transformation processing, contrast normalization processing and whitening processing to obtain low-redundancy sample grey images; 30 deep convolutional neural network model acquiring, wherein the deep convolutional neural network is trainedby means of the low-redundancy training samples by adopting a continuous dropout strategy and tested by means of the test set samples to obtain a deep convolutional neural network model; and 40 pedestrian detection, wherein pedestrian detection is conducted on used field images by means of the deep convolutional neural network model. According to the pedestrian detection method based on the deep convolutional neural network, by means of the continuous dropout strategy, the better generalization capacity can be achieved while the network training precision is kept, and then the higher accuracyrate is achieved in pedestrian detection.

Description

technical field [0001] The invention belongs to the technical field of image recognition, in particular to a pedestrian detection method based on a deep convolutional neural network with fast training convergence speed and strong generalization ability. Background technique [0002] In the field of computer vision, pedestrian detection is a very important research content, and it is widely used in scenarios such as autonomous driving of cars and crowd monitoring in public places. [0003] The traditional pedestrian detection technology is to detect pedestrians by artificially designing features, such as HOG-features, and training classifiers. In the face of scene changes and a sharp increase in the number of detection objects, the cost of artificially designing features is too high to meet real-time requirements and robustness. [0004] Since 2006, deep learning has developed rapidly and has been widely used in areas such as image classification, pattern recognition, and vi...

AI Technical Summary

Problems solved by technology

[0007] To sum up, the problem existing in the existing technology is: when the pedestrian detection is based on the deep convolutional neural network, when the dropout strategy is introduced to enhance the generalization ability of the convolutional network, the trainability of the network may decrease.

Images