基于肤色模型的人耳检测系统

人耳检测是人耳识别以及基于内容的图像和视频检索的一项重要任务。本文提出了一
种将背景差分和肤色模型相结合的人耳检测算法。算法首先在序列图像中自动提取运动目标并进行人体检测，然后经过肤色分割进行人耳的粗定位，产生人耳候选区域。最后利用人耳检测模块判断候选区域中是否含有人耳，以及获得它们的位置、大小等信息。实验结果表明，该算法是有效的。
[关键词] 人耳检测，自适应背景模型，背景差分，肤色模型
[Abstract] Ear detection is an important task in ear recognition and content-based
image and video retrieval. In this paper, we propose a method using background
subtraction and skin color model. First, we drew the movement goal and detected the
human body automatically in the sequence picture, then make a localization roughly and
produce the candidate area of ear via skin color model. Finally, we utilized the module of
ear detection to judge whether to contain the ears of people in the candidate area or not,
and obtained the information of their position, size, etc. Experimental results could
demonstrate its effectiveness.
[key word] ear detection, self-adaptive background update model, background subtraction,
skin clor model