漳州师范学院学报:自然科学版
漳州師範學院學報:自然科學版
장주사범학원학보:자연과학판
Journal of ZhangZhou Teachers College(Natural Science)
2011年
3期
21-24
,共4页
肤色分割%灰度%梯度%几何特征%眼睛定位
膚色分割%灰度%梯度%幾何特徵%眼睛定位
부색분할%회도%제도%궤하특정%안정정위
color segmentation%grey scale%gradient%geometric features%eyes positioning
针对平面旋转人脸的眼睛定位问题,提出基于灰度、梯度及几何特征的眼睛定位方法,在彩色图像中先确定候选人脸区域,分割出眉眼区域,再进行眼睛精确定位及人脸的旋转角度判断.该方法对人脸平面旋转、脸部表情变化等有较好的鲁棒性,且运算量较少,容易实现.
針對平麵鏇轉人臉的眼睛定位問題,提齣基于灰度、梯度及幾何特徵的眼睛定位方法,在綵色圖像中先確定候選人臉區域,分割齣眉眼區域,再進行眼睛精確定位及人臉的鏇轉角度判斷.該方法對人臉平麵鏇轉、臉部錶情變化等有較好的魯棒性,且運算量較少,容易實現.
침대평면선전인검적안정정위문제,제출기우회도、제도급궤하특정적안정정위방법,재채색도상중선학정후선인검구역,분할출미안구역,재진행안정정학정위급인검적선전각도판단.해방법대인검평면선전、검부표정변화등유교호적로봉성,차운산량교소,용역실현.
As for the measures of eyes positioning with plane rotation face,this paper puts forward such methods which are based on grayscale,gradient and geometrical features;the first step is to locate the face field of the person in the colorful pictures,then to segment the eyebrow area,finally to precisely locate the eyes and judge the rotation angle of the face.In this way we can get a better robustness for the face plane rotation and facial expressions,and the less computation makes it easy to implement.