Head area identification method, device and equipment

Abstract

The invention discloses a head area identification method, device and equipment and belongs to the field of machine learning. The method comprises the following steps: obtaining an input image; inputting the input image into cascaded n neural network layers to obtain n groups of candidate identification results of a head area, wherein the neural network layers are used for carrying out identification on the head areas according to preset extraction frames, and at least two neural network layers adopt different sizes of extraction frames; and carrying out aggregation on the n groups of candidate identification results to obtain a final identification result of the head image in the input image. Since the sizes of the extraction frames adopted by at least two neural network layers in the n neural network layers are different, the problem that the head area cannot be identified based on the fixed-size extraction frame when a human face occupies a small area in the monitoring image is solved, so that different sizes of head areas in the input image can be identified, and identification accuracy is improved.

Description

technical field [0001] The present application relates to the field of machine learning, in particular to a method, device and equipment for identifying a human head area. Background technique [0002] Head recognition is a key technology in the monitoring field of public places. Currently, head recognition is mainly done through machine learning models, such as neural network models. [0003] In related technologies, a machine learning model can be used to identify human head regions in surveillance images. The process includes: monitor the image to be tested in areas with large traffic such as elevators, turnstiles, and intersections, and input the image to be tested into the neural network model; the neural network model recognizes image features based on a fixed-size extraction frame , when the image feature matches the face feature, output the analysis result. [0004] Since the head area is identified based on a fixed-size extraction frame, when the area occupied by...

AI Technical Summary

Problems solved by technology

[0005] The embodiment of the present application provides a method, device, and equipment for identifying a human head area, which can solve the problem that the related technology cannot recognize the human face when the area occupied by the human face in the surveillance image is small

Images