文摘
A substantial number of local feature extraction and description methodologies have been proposed as image recognition algorithms. However, these algorithms do not exhibit adequate performance with regard to repeatability, accuracy, and time consumption for both affine transformation and monotonic intensity change. In this paper, we propose a new descriptor, named Resistant to Affine Transformation and Monotonic Intensity Change (RATMIC). Unlike traditional descriptors, we utilize an adaptive division strategy and intensity order to construct the new descriptor, which is actually resistant to affine transformation and monotonic intensity change. Extensive experiments demonstrate the effectiveness and efficiency of the new descriptor compared to existing state-of-the-art descriptors.