Method for counting persons in video based on LSTM (Long Short-Term Memory)-weighted neural network

Abstract

The invention discloses a method for counting persons in a video based on LSTM (Long Short-Term Memory)-weighted neural network, and the method comprises the steps: firstly estimating a perspective according to different scenarios, and then generating an adaptive density map of the crowd; carrying out the down-sampling of continuous multi-frame images, inputting the images to a neural network, andtraining the network based on the images and its corresponding real adaptive density map; estimating the density map of the input images through the trained network, and predicting the number of persons based on the density map. According to the scale difference of the crowd in the scenarios, the method measures the different scale features obtained through network learning through a scale loss function. The method employs a weighted loss function for balancing the contributions of different regions for solving a problem that the crowd is not uniform in distribution. Meanwhile, the method obtains the correlation information of the adjacent frames through LSTM, performs the postprocessing of the predicted number of persons through smooth filtering, and improves the person counting accuracy.

Description

technical field [0001] The invention belongs to the field of intelligent monitoring, in particular to a method for counting crowds on video images by using a long-short-term memory-weighted neural network. Background technique [0002] With the rapid increase of the global population, behaviors such as crowding and stampede in group events continue to increase. On New Year's Day in 2015, 35 people were killed in a vicious stampede on the Bund in Shanghai. Religious activities around the world are also prone to large-scale stampedes, which have extremely serious social impacts. Therefore, monitoring the number of people in real time through surveillance video in public places, and sending an alarm when the number exceeds a certain threshold can effectively avoid vicious group incidents, and can also provide a basis for crowd control. In addition, video-based crowd counting can also be extended to other fields, such as counting the traffic flow on the road and then analyzing ...

AI Technical Summary

Problems solved by technology

However, crowd counting using deep learning still has challenges such as uneven crowd distribution and different scales

At the same time, the current mainstream deep learning crowd counting methods usually only predict the number of people in still images, but cannot predict the number of people in video images with rich temporal information.

Images