Illumination-insensitive Binary Descriptor for Visual Measurement Based on Local Inter-patch Invariance

Binary feature descriptors have been widely used in various visual measurement tasks, particularly those with limited computing resources and storage capacities. Existing binary may not perform well for long-term tasks due to their sensitivity illumination variations. It can be observed that when image changes dramatically, the relative relationship among local patches mostly remains intact. Based on observation, consequently, this study presents an illumination-insensitive (IIB) descriptor by leveraging inter-patch invariance exhibited multiple spatial granularities deal unfavorable By taking advantage of integral images patch computation, a highly efficient IIB is achieved. encode scalable features granularities, thus facilitating computationally hierarchical matching from coarse fine. Moreover, also apply other types data, such as depth maps semantic segmentation results, available some applications. Numerical experiments both natural synthetic datasets reveal proposed outperforms state-of-the-art testing float descriptors. The has successfully employed demo system localization. code 1 will publicly available.