Human eyes location using wavelet and neural network
Yanjiang Wang, Baozong Yuan · 2002
This paper presents a new method for the detection and localization of human eyes in a color image. First, the face-like regions are located using an evolutionary color clustering algorithm. Then, wavelet decomposition is applied to each region to detect the possible facial features. If the features are distributed like a predefined human face model, the eyes can be located; otherwise, a three-layer BP neural network is used to check if there is an eye in the region. Experimental results show that our method is fast and has a good performance.