An Improved ORB Feature Point Extraction Algorithm based on Image Enhancement and Adaptive Threshold
Tao Jiang, Yuchen Xue · 2023
In scenes with varying illumination, there can be poor quality of the captured images, resulting in feature points not being extracted or not being extracted in sufficient quantity when using the fixed-threshold ORB algorithm to extract feature points. To address this problem, an improved ORB feature point extraction algorithm is proposed. The image quality is improved using the adaptive gamma transform. According to the greyscale distribution of the image, the brightness change of the image is judged and the image is divided into three types of images: no obvious brightness change overall, brightness change overall and local brightness change. The feature point extraction is completed by adopting a global adaptive threshold or global adaptive threshold combined with a local adaptive threshold for different image types respectively. The experimental results show that the improved ORB feature point extraction algorithm can significantly increase the number of image feature points extracted with good robustness in lighting change scenes.