【摘要】 针对现有数字图像运动估计算法不能完全适用于低分辨率触觉图像的问题，提出了一种自适应参考模板互相关(ARTC)的触觉图像位移估计算法，其流程包括模板自适应选取、最相似子区搜索和亚像素位移估计。该算法自适应选取模板以实现触觉图像位移的快速、高精度估计。实验结果表明：本文提出的ARTC算法对低分辨率触觉图像进行位移估计的精度和耗时都优于相位相关算法、仿射法和尺度不变特征变换(SIFT)算法。更多还原

【Abstract】 Aiming at the problem that existing digital image motion estimation algorithms are not fully applicable to low-resolution tactile images, an adaptive reference template cross-correlation(ARTC)algorithm is proposed for displacement estimation of tactile images.Its process includes template adaptive selection, most similar sub-region search, and sub-pixel displacement estimation.The algorithm adaptively selects templates to achieve fast and highly accurate estimation of tactile image displacement.The experimental results show that the proposed ARTC algorithm have better performances in precision and time consumption than those of the phase correlation algorithm, affine method, and scale-invariant feature transform(SIFT)algorithm for displacement estimation of low-resolution tactile images.更多还原