运动视觉误差分析的模型设计与实现

Design and implementation of motion vision error analysis model

针对运动目标模型三维建模中,视觉误差导致建模的准确性不好的问题,提出一种基于运动视觉误差补偿的三维模型设计方法。采用全息投影方法进行三维运动模型图形采集,结合边缘像素特征提取方法进行三维模型的体绘制,根据运动视觉误差补偿方法构建图像重构的三维数据场,进行运动模型的边缘轮廓特征提取和Harris角点检测,实现运动模型的三维重建和视景仿真重构。仿真结果表明,采用该方法进行运动模型重建,输出三维模型的噪点较小,轮廓特征配准精度较高,运动图像识别的视觉误差收敛到零,提高成像质量。

Since the modeling accuracy is poor due to vision error in 3D modeling of moving object model, a 3D model de- sign method based on motion vision error compensation is put forward. The holographic projection method is used to perform graph collection of 3D motion model, and combined with the edge pixel feature extraction method to carry out volume rendering of 3D model. 3D data field of image reconstruction is constructed according to motion vision error compensation method to ex- tract the edge contour feature of motion model and detect the Harris angular point, so as to realize 3D reconstruction of motion model and visual simulation reconstitution. The simulation results show that the method used to reconstruct the motion model can output small noisy point of 3D model, has high registration accuracy of contour features, and the vision error of moving image recognition can converge to zero, which can improve the imaging quality.