People number counting method combined with convolution neural network and track prediction

Abstract

The invention relates to a people number counting method combined with convolution neural network and track prediction. The method comprises following steps of using the frame difference method to segment crowd blocks included in a video; distinguishing a sparse crowd block mass and a dense crowd block mass; for the sparse crowd block mass, converting an RGB image through a formula so as to obtainHSV color spaces; in two different color spaces, using a selection searching algorithm to pre-determine pedestrian positions; combining and removing repeated regions in the two spaces so as to obtainpedestrian region positions; using the convolution neural network to extract characteristics and selecting grid loss function Grid Loss to train the network based on the blocks, thereby achieving recognition of local positions including faces and bodies of shielded pedestrians; for the dense crowd block mass, extracting characteristics of a crowd density distribution graph, establishing a multi-regression model and estimating people number; and using the Markov model chain to predict walk tracks of the shielded pedestrians, locking the positions of the shielded pedestrian and counting the pedestrians.

Description

technical field [0001] The present invention relates to the field of target recognition, in particular to a method for counting people in combination with convolutional neural network and trajectory prediction. The method is mainly used in video people counting represented by deep learning and trajectory prediction for target detection. Background technique [0002] People counting is a research direction with practical significance in the field of intelligent video surveillance in recent years. It is mainly divided into people counting methods based on target detection and people counting methods based on feature regression. Both types of methods use supervised machine learning. In addition, there are tracking trajectory clustering methods based on unsupervised learning. In supervised learning methods, pedestrian detection based on the HOG algorithm is one of the widely used methods. This method constitutes pedestrian features by calculating and counting the gradient direct...

AI Technical Summary

Problems solved by technology

Although these algorithms have extracted effective and high-quality pedestrian features and established accurate prediction models, they still cannot solve the impact of factors such as occlusion, illumination changes, and uneven crowd distribution on detection.

Images