The eyes can be considered as a salient and relatively stable feature of faces , so firstly eyes - analogue regions in cluttered images are segmented using local adaptive threshold edge detector . then the small eye - analogue regions are grouped together and labeled using a traditional labeling process according to their geometrical and color features . instead of finding potential eye - pairs from eye - analogue regions one by one , all possible pairs of eyes are encode as the solution , and the potential face are searched by gas 新意之處:利用膚色是一種感知現(xiàn)象的特性以及神經(jīng)網(wǎng)絡(luò)的強(qiáng)大感知、學(xué)習(xí)功能,采用進(jìn)化神經(jīng)網(wǎng)絡(luò)對(duì)膚色進(jìn)行感知和分類;以眼睛的幾何和周邊的膚色特征分割眼睛塊,克服了在完整的膚色區(qū)域內(nèi)檢測(cè)人臉?lè)椒ǖ木窒扌?;?duì)候選眼睛對(duì)的編號(hào)而不是人臉在圖像中的位置、人臉的尺度和方向進(jìn)行編碼,使遺傳算法用于人臉定位時(shí)的問(wèn)題空間得以大大簡(jiǎn)化,采用一個(gè)新穎的眼睛及臉頰部位的灰度投影及膚色模板驗(yàn)證,可以實(shí)現(xiàn)多尺度、任意方向的正面人臉檢測(cè),實(shí)驗(yàn)結(jié)果令人滿意。