论文部分内容阅读
Automatic and robust matching of multi-modal images can be challenging owing to the nonlinear intensity differences caused by radiometric variations among these images. To address this problem, a novel dense feature descriptor and improved similarity m