Object Recognition Using a Multichannel Vision System

1.利用多通道图像估计物体的反射模型提出了一种利用物体表面的图像数据和距离数据来估计反射模型的各种参数的方法。建立了激光测距与多波段摄像系统相结合的统一测量系统，获取物体表面的三维面形数据和光谱反射率数据。首先，利用两个光照方向上的观测图像和根据距离数据计算的表面法向向量，获得每个像素处的漫反射分量和镜面反射分量。然后根据漫反射分量估计光谱反射率。然后，将提取的镜面反射分量与托兰斯-斯帕罗模型的镜面反射函数进行拟合。2.利用多通道视觉系统测量和绘制艺术画我们提出了一种利用RGBCome…的简单成像系统来测量艺术画的方法更多的ra和光源，并在任何照明和观察条件下将绘画渲染为CG图像。表面形状被认为是在平坦的画布上涂上油漆的粗糙平面。我们捕捉了同一幅画在不同光照方向下的一些图像。首先，基于光度学立体，从漫反射分量估计曲面法线。接下来，根据漫反射组件的RGB值估计曲面光谱反射率。3.基于多通道视觉系统的印制电路板识别提出了一种基于表面光谱反射率的原始电路板目标材料分类方法。该多光谱成像系统由液晶可调谐滤光片、单色CCD摄像机、微距镜头和个人计算机组成。我们已经证明，像塑料这样的介电材料可以根据变化的照明几何形状的反射比差异而与金属区分开来。在此基础上，提出了一种将目标划分为多个电路元件的算法。电路板的区域分割结果在实际电路板上进行了实验验证。较少

1.Estimation of reflection models of objects by using multichannel imagesA method is proposed for estimating various parameters of a reflection model using both the image data and the range data of an object surface. A unified measuring system, combining a laser range and a multi-band camera system, is made for acquiring the 3D shape data and the spectral reflectance data of the object surface. First, the diffuse reflection component and the specular reflection component at every pixel are obtained using the observed images at two illumination directions and the surface normal vectors calculated from the range data. The spectral reflectance is then estimated from the diffuse reflection component. Next, the extracted specular reflection component is fitted to the specular function of the Torrance-Sparrow model.2.Measuring and rendering of art paintings by using a multichannel vision systemWe propose a method for measuring an art painting by using the simple imaging system of an RGB came … More ra and a light source, and rendering the painting as CG images under any conditions of illumination and viewing. The surface shape is considered as the rough plane with paints on a flat canvas. We capture some images of the same painting under different illumination directions. First, the surface normal is estimated from the diffuse component, based on a photometric stereo. Next, the surface-spectral reflectance is estimated from the RGB values for the diffuse component. Moreover, the Torrance-Sparrow model is used for image rendering.3.Recognition of printed circuit boards by using a multichannel vision systemWe propose a method for classifying object materials on a raw circuit board based on surface-spectral reflectance. The multi-spectral imaging system is composed of a liquid-crystal tunable filter, a monochrome CCD camera, macro-lens and a personal computer. We have shown that dielectric materials like plastics can be distinguished from metals based on the reflectance difference in changing illumination geometries. Then an algorithm is presented for classifying the objects into several circuit elements. Region segmentation results of the circuit board are demonstrated in an experiment using a real board. Less