Study on an environmental embedding type "eye camera"

在这项研究中，我们提出了一个复杂的方法来估计视觉方向虹膜轮廓。这种方法只需要一个未知焦距的相机拍摄的单目图像。为了估计视觉方向，我们假设双眼的视觉方向是平行的，虹膜边界是3D空间中的圆形。在这种情况下，虹膜边界所在的两个平面也平行。我们利用一种扩展的“两圆”算法从输入图像中提取虹膜轮廓来估计这两个平面的法向量，并提出了一种通用的目标跟踪方法（称为“K-均值跟踪器”），该方法通过基于K-均值聚类的算法将搜索区域内的像素分为“目标”和“背景”两组来实现。我们首先使用5D特征向量来统一描述每个像素的颜色（“Y，U，V”）和位置（“x，y”）。这使得能够在跟踪期间同时适应颜色和几何特征。其次，我们使用一个可变的椭圆模型来描述搜索区域的形状和周围的背景模型。这保证了稳定的目标跟踪下的各种几何transformation. Further，我们提出了一个高性能的目标跟踪系统，用于获得高质量的图像的运动物体在视频速率通过控制安装在电机驱动的云台单元上的主动摄像机。我们使用“K-均值跟踪器”来跟踪图像空间上的对象。采用PID控制方案，根据“K-均值跟踪器”的结果来控制云台单元的角位置和速度。利用该系统，我们可以获得运动物体的无运动模糊图像。

In this research, we propose a sophisticated method to estimate visual direction using iris contours. This method requires only one monocular image taken by a camera with unknown focal length. In order to estimate the visual direction, we assume the visual directions of both eyes are parallel and iris boundaries are circles in 3D space. In this case, the two planes where the iris boundaries reside are also parallel. We estimate the normal vector of the two planes from the iris contours extracted from an input image by using an extended "two-circle" algorithm.In this research, we also propose a general object tracking method (called as "K-means tracker"), which is realized by classifying the pixels within the search area into "target" and "background" groups with a K-means clustering based algorithm. We first use a 5D feature vector to describe both the color ("Y,U,V") and the position ("x,y") of each pixel uniformly. This enables the simultaneous adaptation to both the color and geometric features during tracking. Secondly, we use a variable ellipse model to describe the shape of the search area and to model the surrounding background. This guarantees the stable object tracking under various geometric transformations.Moreover, we propose a high performance object tracking system for obtaining high quality images of a moving object at video rate by controlling an active camera mounted on a motor droved pan-tilt unit. We use "K-means tracker" for tracking object on image space. The PID control scheme is employed for controlling the angular position and speed of the pan-tilt units according to the results of the "K-means tracker". With this system, we can obtain motion-blur-free images of a moving object.