A Portrait Localization Method Based on Local 2D Pattern and Invariant Moment Search

Abstract

The invention discloses a method for locating portraits based on local two-dimensional patterns and invariant moment search. The method includes: obtaining training results of portrait samples, standardized LBP eigenvalues ​​and portrait outer contour features of Hu's invariant moments, and The shape of the outer contour is similar to an ellipse, and Hoo's invariant moment is used to collect the geometric features in the image area to obtain the contour feature value of the portrait in the image; by first performing color gamut transformation on the image containing the portrait, it is converted to gray degree map, and then perform filtering processing to remove the interference of color and noise, and leave an image with obvious texture features; then perform feature analysis of the local binary pattern (LBP) to obtain the LBP feature value of the edge in the image, through this The LBP eigenvalue further combines the contour eigenvalue of the invariant moment to distinguish the foreground and background of the portrait in the image, so as to find the exact position of the portrait in the complex background environment and achieve the purpose of fast search and positioning.

Description

technical field [0001] The invention relates to various image processing technical fields such as image segmentation and edge search, and also has certain requirements for signal filtering, and is generally a portrait positioning method based on local two-dimensional patterns and invariant moment search. Background technique [0002] Face recognition is now a very challenging research, and it has received more attention because of its wide range of applications. Face recognition technology has countless applications in business, military, intelligence and daily life. Simply put, there are information security management, medical care, security systems, artificial intelligence, case investigation and so on. The most important ones are the following three applications: 1. Identity authentication. By comparing the portrait data collected in real time with the portrait data stored in the authentication device, if they reach a certain degree of similarity, it can be judged that ...

AI Technical Summary

Problems solved by technology

Image segmentation is one of the most basic and important fields in low-level vision in the field of image processing and computer vision. It is the basic premise for visual analysis and pattern recognition of images. At the same time, it is also a classic problem. So far, neither A common image segmentation method, there is no objective standard for judging whether the segmentation is successful

Images