Facial expression capturing method and device in weak light environment

Abstract

The invention provides a method and a device for capturing facial expressions in a weak light environment, which can realize accurate framing of seats with different scales in a video stream in a complex environment, and detect and extract facial features for identifying expressions based on the framing result. The method is characterized by comprising the following steps: S1, acquiring a video stream of an overlook angle in a cinema to be detected; s2, detecting a video stream based on a cinema seat positioning model to obtain a labeling box of the newest positions of all seats in a cinema to be detected; and S3, performing face positioning and expression recognition on all the labeling boxes by using an expression recognition classification model so as to obtain face expression categories of audiences in the to-be-detected cinema. Wherein the cinema seat positioning model takes a network combined by ResNet50 and FPN as a Backbone, a CBAM module is added to construct a cinema seat positioning initial network, and the initial network is trained to obtain the cinema seat positioning model, and the expression recognition classification model is obtained on the basis of constructing and training a facial expression recognition algorithm combined by improved VGGNet and improved Focaal Loss.

Description

technical field [0001] The invention belongs to the field of computer vision, and in particular relates to a method and device for capturing human facial expressions in low-light environments. Background technique [0002] Scientific statistical methods of movie viewing data and intelligent methods of maintaining movie viewing order are indispensable links in the modernization of the film industry. Box office data is an important criterion for movie revenue statistics, and it is an intuitive representation of whether a movie is successful or not. The real attendance rate, audience status during the screening, such as facial expressions, and the number of departure times, are used to evaluate audience viewing from more dimensions. An important basis for the movie experience, which provides an important reference for the film producer to control the rhythm of the film and provide a better viewing environment for the screening party; maintaining a good viewing order is the resp...

AI Technical Summary

Problems solved by technology

However, the facial expression recognition method does not achieve good results in low light conditions, so it is necessary to improve the existing methods

Images