基于改进ORB的抗视角变换快速图像匹配算法

针对快速定向二进制描述（ORB）算法误点率高、像素点对对比易受噪声影响同时抗视角变换能力弱等问题，本文提出基于改进ORB的抗视角变换快速图像匹配算法。该算法在ORB算法的架构下利用加速分割检测算法进行特征提取，并提出特征点加速检测策略，利用改进的二进制鲁棒独立基本特征算法进行特征点描述，同时结合透视变换模型进行图像采样和汉明距离作为相似度量准则进行图像匹配，初匹配后利用网格运动统计算法去除误匹配点。实验表明，本文算法在保持实时性的同时，能够获得较高的匹配精度，和较强的抗视角变换能力，具备较好的工程意义。

Anti-perceptual Transform Fast Image Matching Algorithm Based on Improved ORB

Aiming at the problem of oriented fast and rotated brief (ORB) algorithm,such as high false matching rate,pixel points contrastive are susceptible to noise and weak ability of anti-perceptual transformation,this paper proposes anti-perceptual transform fast image matching algorithm based on improved ORB.The algorithm uses the adaptive and generic accelerated segment test algorithm to extract features under the framework of the ORB algorithm,and proposes a feature point acceleration detection strategy.Then the improved binary robust independent elementary features algorithm is used to describe the feature points.At the same time,image sampling combined with perspective transformation model and Hamming distance as similar metrics for image matching. After the initial image matching, the grid-based motion statistics algorithm is used to remove the false matching points.The experiment shows that the proposed algorithm can achieve high matching accuracy and strong ability to anti-perceptual transformation while maintaining real-time performance, and it has a good engineering significance.