An End-to-End Convolutional Neural Network Pedestrian Detection Method

Abstract

The invention discloses an end-to-end convolutional neural network pedestrian detection method, which is used to solve the problems of the existing pedestrian detection algorithm such as low detection accuracy, complex algorithm and difficulty in multi-module fusion. A new end-to-end convolutional neural network is adopted. By constructing a labeled training sample set and using end-to-end training, a convolutional neural network model that can predict the confidence of the pedestrian candidate frame and the corresponding frame is obtained. During the test, input the test picture into the trained model to obtain the corresponding pedestrian detection frame and confidence. Finally, non-maximum suppression and threshold screening are performed to obtain the best pedestrian area. Compared with previous inventions, the present invention has two advantages: one is end-to-end training and testing, which makes the training and testing of the entire model extremely easy; the other is that the present invention solves the problem of pedestrian scale and proportion by constructing a candidate frame regression network , does not need the pyramid technology used in previous inventions, which greatly saves computing resources.

Description

technical field [0001] The invention belongs to the technical field of pattern recognition, in particular to an end-to-end convolutional neural network pedestrian detection method. Background technique [0002] Pattern recognition is an important interdisciplinary subject in the field of artificial intelligence and image processing, and it is a research hotspot in recent years. Early task solutions mainly relied on human professional domain knowledge to design an algorithm or build a system, and the problem-solving effect was limited by human knowledge. With the development of artificial intelligence, the convolutional neural network was formally proposed by Lecun in the 1990s. Hinton improved the original convolutional neural network in 2012 and achieved the first result in the ImageNet competition. Since then, convolutional neural networks have been widely used in fields such as computer vision, natural language processing, and intelligent search. Convolutional neural n...

AI Technical Summary

Problems solved by technology

[0003] After searching the existing technology, it was found that the Chinese Patent Document No. CN105335716A was published (announced) on 2016.02.17, which disclosed a pedestrian detection method based on improved UDN to extract joint features, including: image preprocessing; The preprocessed image extracts the overall and local features of the human body; weights the category probabilities output by the overall features and local features of the step to obtain the final probability value, and judges whether the original input image contains pedestrians according to the final probability value. Compared with the low missed detection rate, the actual missed detection rate is still high, and the technology has not achieved end-to-end training, and must rely on the HOG+CSS++SVM algorithm to obtain the initial candidate area for pedestrians

This technology requires module cascading, which makes the whole system more complicated

Images