Crowd counting method based on multi-scale space guide perception aggregation network

Abstract

The invention discloses a crowd counting method based on a multi-scale space guide perception aggregation network. The crowd counting method comprises the following steps: S1, establishing a multi-scale feature extraction network; s2, inputting the image with any resolution into a multi-scale feature extraction network; s3, inputting the multi-scale features captured by the multi-scale feature extraction network into a space guidance network, and outputting context guidance perception features and a guidance perception graph; s4, transmitting the context guide perception features and the guide perception map to an attention fusion network, finally outputting a density map, and constructing a density map training set; s5, establishing an adaptive scale loss function, and performing adaptive training through the density map training set; and S6, taking the to-be-predicted image as input, repeating the steps S2-S5, and outputting a crowd calculation result in the to-be-predicted image. Multi-scale information aggregation is carried out on a self-adaptively captured space environment by utilizing a reasonable and efficient guiding method, and the counting accuracy and robustness are improved.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a crowd counting method based on a multi-scale space-guided perceptual aggregation network. Background technique [0002] With the increasing concentration of the global urbanized population, computer vision-based crowd counting and recognition technologies play an important role in public safety, abnormal event detection, and urban traffic management. In sparse scenes containing single or multiple objects in the image, the object localization detection technology can easily and accurately perform crowd counting and identification. Due to the gradual development of deep learning, methods based on convolutional neural networks (CNN) have achieved remarkable success in tasks such as image classification, pedestrian detection, and speech recognition. Therefore, researchers introduced CNN into the field of crowd counting, and achieved good crowd density estimation results in...

AI Technical Summary

Problems solved by technology

However, capturing detailed features is not enough and suffers from underperception in this task

[0006] 3. Complex background interference is also an important factor affecting the accuracy of crowd counting. The previous model solved the above problems through self-attention mechanism or image segmentation, but the complexity of the model is high and the calculation cost is relatively high.

[0007] 4. In feature fusion, simple fusion will weaken the effectiveness of information dissemination. In the field of crowd counting, especially in the field of spatial location, advanced semantic information and background interference.

Images