论文部分内容阅读
用神经网络方法解决未知光源位置环境中物体三维形状恢复的问题．对漫反射表面，用神经网络方法由已知表面形状物体及其对应图像的灰度值进行学习，所得权值可视为环境光源参数．由此可恢复同样光源环境中其它物体的三维形状．实验证明，神经网络方法可以解决未知光源位置环境（包括多个光源）中漫反射表面物体的三维形状恢复问题．
Using neural network method to solve the problem of three - dimensional shape recovery of objects in unknown light source position. For the diffuse surface, the neural network method is used to study the grayscale value of the surface-shaped object and its corresponding image. The weight can be regarded as the ambient light source parameter. This restores the three-dimensional shape of other objects in the same light environment. Experiments show that the neural network method can solve the problem of three-dimensional shape restoration of diffusely reflecting surface objects in the unknown light source environment (including multiple light sources).