Multi-illumination face recognition method based on morphologic quotient images

Abstract

The invention provides a multi-illumination face recognition method based on morphologic quotient images. The method comprises the following steps that: illumination intensity of each position of a face is estimated, including light-source intensity and normal vector information; a quotient value of an original illumination image and an illumination estimation image is evaluated so as to obtain the reflectivity information of the face surface, namely face texture features unrelated to illumination conditions; because dividing operation can produce noise points which are unrelated to the face texture features in the prior shadow region, the original image is preprocessed before illumination estimation; illumination estimation adopts a morphologic closed operation method and uses the simplest and most convenient rectangular mean template; the scale size of the template adopts a dynamic principle; according to the characteristics of different local regions, the size of the template is chosen through self-adaptation; a quotient image result obtained through division is subjected to nearest neighbor classification; and a distance criterion adopts normalized correlation. The method has the advantage of improving the security of automatic face recognition systems, and has important application value in the field of biometrics recognition.

Description

technical field [0001] The invention relates to the technical field of biological feature recognition, in particular to using morphology to process human face images under different lighting conditions. Background technique [0002] Because face recognition is in line with people's habit of identifying in life, and the face has certain stability and reliability, the face image is easy to obtain, and the equipment cost is low. Face recognition technology has become a kind of technology in modern society. important means of identification. Face recognition technology is widely used in security systems such as banks, customs, public security, and home security equipment. In addition, due to the advantage of non-contact in the collection of face images, it has been widely used in video surveillance. [0003] However, although with the advancement of technology, the performance of the face recognition system has achieved a high readiness rate in an ideal environment, various ex...

AI Technical Summary

Problems solved by technology

The biggest disadvantage of this method is that the calculation time complexity is very high, and it is not suitable for application in real-time face recognition systems.

[0006] Based on the above analysis and research, it is found that the existing face recognition algorithms in the world have not really solved the problem of uneven illumination quickly and effectively.

Images