Aiming at the problem that how to extract stable feature points from low illumination images

a feature point detection method combining image logarithmic domain transform with traditional detection method is proposed.Aiming at the logarithmic domain transformation

an adaptive basis is introduced

and different basis are used for different information entropy.Furthermore

the detection effect of low-illumination image under different illumination degree and the illumination robustness of the algorithm are analyzed.In the experimental part

it is verified that the recommendation algorithm has a great improvement in information entropy

matching point pairs and repetition rate compared with the traditional feature point detection method

which proves that the method can effectively improve the feature detection performance of low-illumination images.