Discrimination of objects Based on Surface Reflectance Characteristics

1.非均匀物体的颜色反射模型非均匀材料的标准反射模型假定光通过两个独立的机制反射。我们开发了一种测试标准反射模型的方法。其次，提出了一种从两个物体表面反射的观测光谱估计光源光谱分布的计算方法。通过对一些非均匀塑性物体的实验，验证了模型的有效性和方法的可靠性。各种物体的颜色反射模型基于对物体表面反射光谱及其表面光谱反射率的分析，提出了各种物体的颜色反射模型。我们已经提出了一种测试标准反射模型在日常生活中对真实对象的充分性的过程。测试结果表明，标准模型适用于多种材料，但不适用于只由镜面反射组成的金属，也不适用于某些纸张和布料。通过对这些材料的表面反射率的详细分析，表明光反射可以用扩展的双色反射模型来表示。基于标准反射模型的表面识别许多物体表面被假定为由非均匀材料组成，其表面光谱反射函数被描述为镜面反射和漫反射之和。镜面反射分量被用来估计光源的光谱功率分布，而漫反射分量被用作表面一致性的主要指示。在场景的实际测量中，从不同的空间位置观察到的表面光谱反射率的数量取决于不同的身体成分。在这里，我们提出了一种分两步恢复每个表面漫反射函数的方法。

1. A color reflection model for inhomogeneous objectsThe standard reflection model for inhomogeneous materials assumes that light is reflected by two independent mechanisms. We have developed a method for testing the standard reflection model. Next a computational method has been proposed for estimating the illuminant spectral distribution from observed spectra reflected from two object surfaces. The validity of the model and the reliability of the method are confirmed on experiments using some inhomogeneous objects of plastics.2. Color reflection models for a variety of objectsColor reflection models have been proposed for a variety of objects based on the analysis of the reflected light spectra from the object-surfaces and their surface spectral reflectances. We have presented a procedure for testing the adequacy of the standard reflection model on real objects in our everyday lives. The test results shows that the standard model is adequate for many materials, but inadequate for metals composed of only specular reflection, and for some of papers and cloths. By a detailed analysis of the surface reflectance of these materials, it has been shown that the light reflection can be expressed in an extended dichromatic reflection model.3. Surface identification using the standard reflection modelMany surfaces of objects are assumed to be composed of inhomogeneous materials, whose surface spectral reflectance function is described as the sum of a specular reflectance and a diffuse reflectance. The specular component is used to estimate the spectral power distribution of the illuminant, while the diffuse component is used as the principal indication of the surface identity. In real measurement of a scene, a number of the surface-spectral reflectances observed from spatially different locations depends on the different body components. Here we have developed a method for recovering the diffuse reflectance function of each surface in two steps.