Classic Visual Feature Descriptors
Experiments show that different visual descriptors tend to have similar rates of outliers in feature matching, and the precision differences brought by the application of different feature descriptors in Visual SLAM are negligible.
