图像特征点提取角点量阈值自适应选取方法

Image Point Feature Extraction Based on Adaptive Corner Interesting Value Threshold Choosing Algorithm

针对图像特征点提取中在固定的角点量阈值时,由于拍摄物体自身纹理受光照的不均匀,可能会引起角点的提取效果不佳的现状,提出一种角点量自适应选取方法。阐述了角点量自适应选取方法的增点法和删点法两种实现形式,并分别通过对模拟和实拍图像特征点的提取,定量和定性验证了自适应角点量阈值方法相对于固定角点量阈值方法的优势。

Aiming at the dissatisfied feature points extracting result caused by the uneven illumination on the measured object when the corner interesting value threshold is given before image processing, a new adaptive corner interesting value choo- sing method is put forward. Two different realization form of this new method, adding point method and deleting point method are explained in detail. Both images of simulation and real taken are used to validate the advantage of adaptive corner interesting value threshold choosing method qualitatively and quantificationally.