A Face Image Detection and Filtering Method

Abstract

The invention relates to a face image detection and filtering method, which is optimized and designed for the existing face detection image technology, through the difference between the real face image and the cartoon face image, in terms of pixel color mean and variance, training and obtaining Identify the filter classifier of real face images and cartoon face images, and then realize the classification of face detection images through the obtained filter classifier, filter out the cartoon face images, and realize the filtering of face detection images. Real face images are obtained, and the entire design method process effectively improves the accuracy and reliability of existing face detection technologies.

Description

technical field [0001] The invention relates to a face image detection and filtering method. Background technique [0002] In the Internet society, there is a large amount of multimedia data on the network, and image data is overwhelming in the network society. There is a large amount of visual information in the image, which can reflect the information to the viewer. However, the current digital image and photo coding form is all based on red, green and blue pixels. Composition, like computer text data, it is not easy to read objects (such as faces, vehicles) on photos in the form of direct string matching. When there is a face in the photo, it is necessary to use human eyes to sense Therefore, there are products on the market specifically for face detection in pictures, which can locate the face area on a digital image. [0003] Although some face detection products have been launched on the market, these products have certain deficiencies. The current face detection tech...

AI Technical Summary

Problems solved by technology

[0002] In the Internet society, there is a large amount of multimedia data on the network, and image data is overwhelming in the network society. There is a large amount of visual information in the image, which can reflect the information to the viewer. However, the current digital image and photo coding form is all based on red, green and blue pixels. Composition, like computer text data, it is not easy to read objects (such as faces, vehicles) on photos in the form of direct string matching. When there is a face in the photo, it is necessary to use human eyes to sense Therefore, there are products on the market specifically for face detection in pictures, which can locate the face area on a digital image

[0003] Although some face detection products have been launched on the market, these products have certain deficiencies. The current face detection technology can locate areas similar to human faces on the image, but there are many problems with the authenticity of detected faces. The problem is that some cartoon face pictures, cartoon expression pictures, etc. will be detected as real faces, but in the actual process, people need to detect and recognize real faces for subsequent analysis.

Images