可信度计算是深度图融合三维重建中非常重要的一步,直接关系到重建结果的完整性和精度。提出一种充分考虑图像尺度、三维点的几何特性、三维点与摄像机之间的关系以及图像匹配程度等因素的可信度计算方法,可信度计算更精确。实验结果表明,该方法能提高深度图融合三维重建结果的完整性和精度。
Confidence computing is a very important step in depth map fusion based three-dimensional reconstruction which is directly related to the completeness and accuracy of the reconstruction results. Proposes a method that is fully considered image scale, geometry orientation, the angle between the point and cameras, image match cost and other factors, the confidence calculated is more accurate. Experimental re-sults show that the method can improve the depth map fusion based three-dimensional reconstruction results in the completeness and ac-curacy.