Pedestrian recognition method based on positive-negative generalized max-pooling

Abstract

The invention discloses a pedestrian recognition method based on positive-negative generalized max-pooling. The method comprises: preprocessing acquired traffic videos to obtain required training sample images, extracting local features of the training sample images by means of gradient-based HOG local descriptors, encoding the local features by the depth hierarchical encoding method which is completed via a spatial aggregating restricted Boltzmann machine, forming feature encoding vectors of training samples, obtaining high-level image feature representation vectors by adopting the positive-negative generalized max-pooling method, and inputting feature data obtained to a support vector machine classifier to finish the training; preprocessing to-be-tested pedestrian images to obtain test samples, and obtaining feature representation vectors of the test samples by the same method; inputting the feature representation vectors of the test samples to the support vector machine classifier which is already trained, and identifying whether test images are pedestrian images or not. The invention improves the accuracy of pedestrian recognition and enhances the robustness of the pedestrian recognition algorithm.

Description

technical field [0001] The invention belongs to the technical field of computer vision, and in particular relates to a pedestrian recognition method based on positive and negative generalized maximum pooling. Background technique [0002] Pedestrian recognition has broad application prospects in intelligent transportation systems and intelligent monitoring systems, but it is still an open problem in the field of computer vision, because pedestrian appearance and background environment, such as clothing, posture, lighting, viewing angle, etc., vary greatly. Coupled with the complex background, the recognition accuracy is not high. [0003] Feature pooling is becoming more and more important in the entire pedestrian detection system. The idea of ​​feature pooling comes from the study of complex cells in the striate cortex, and it has become a common method for image / video feature representation and encoding. The basic pooling methods for pedestrian recognition are max poolin...

AI Technical Summary

Problems solved by technology

Sum pooling is suitable for any encoding, however, due to the fact that non-informative descriptors frequently affect the result representation, while potential highly informative descriptors have little effect on the result representation, sum pooling lacks discriminative power

Average pooling is a quantization method that considers all elements of the pooled area in average pooling. However, this method easily loses the spatial information of each block.

l p Pooling has ideal discriminative pooling results, l p Pooling is to model it in a more exhaustive way, but l p The pooling stage algorithm needs further research

Images