Uses blender to generate a scene with reference objects. Takes input images and uses opencv methods to estimate the object translation and rotation from the camera. Reads the object 3D coordinates from a file and uses color to detect the objects in t

利用opencv自带的函数功能，编译出由两个摄像机，既双目视觉来获得物体的深度信息，最终得到三位映射。-Opencv own function, compile the object depth information obtained by two cameras, binocular vision, and ultimately get three mapping.

由双目立体视觉系统匹配到的点来计算物体的真实坐标，对三维物体进行三维重建。-The real coordinate by binocular stereo vision system matching to the point to calculate the object, 3D reconstruction of 3D objects.