Face recognition preprocessing method based on grey-scale map boundary detection and noise frame filling

Abstract

The invention discloses a face recognition preprocessing method based on grey-scale map boundary detection and noise frame filling. According to the method, the area contour of the human face is recognized through gray image boundary detection of the human face image, and 0-value filling is performed on the areas outside the main feature area of the human face according to the human face contour principle, so that frame interference factors including hair in the areas outside the human face are removed, and the preprocessing precision of human face cutting is further improved. The face recognition preprocessing method based on grey-scale map boundary detection and noise frame filling can further reduce interference factors of face recognition and improve the accuracy of face recognition, and has the advantages of being high in face processing accuracy, reducing face recognition interference and improving the accuracy of face recognition.

Description

technical field [0001] The present invention relates to the technical field of face recognition, in particular to a face recognition preprocessing method for grayscale image boundary detection and noise frame filling with high face processing accuracy, reduced face recognition interference, and improved face recognition accuracy. Background technique [0002] With the development of society and the advancement of science and technology, artificial intelligence technology has also developed by leaps and bounds. Various artificial intelligence technologies are applied to all aspects of our lives, bringing many conveniences and surprises to people's lives. In the near future, Artificial intelligence technology is bound to bring revolutionary changes to people's production and life. Especially in the field of intelligent security, the implementation of application fields such as face access control, license plate recognition, sound equipment and voice wake-up and related technol...

AI Technical Summary

Problems solved by technology

However, any artificial intelligence algorithm has a certain false positive and recognition rate, and the face recognition algorithm is no exception

In terms of face recognition, face alignment and cropping are one of the key steps. How to align and crop images to maximize the inclusion of face information and minimize the acquired image matrix is ​​the key to improving this step. In the image matrix, there will still be interference factors other than the face, such as the background frame and hair around the head portrait, which are often changing and have great interference factors for the calculation of face features.

Images