Laser radar invariant spatial chromatic image descriptor

Abstract

Laser radar (LADAR) is a three-dimensional (3-D) imaging system which provides very high-resolution 3-D images and is used for automatic recognition of 3-D targets. In order to achieve a high recognition ability for this system, a new LADAR image descriptor is proposed. This descriptor extracts features from the LADAR images by using special chromatic processors called invariant half-height overlapping triangular processors. The descriptor developed has a high discrimination capability and is robust to the effects that disturb LADAR images. The performance of the proposed LADAR descriptor is evaluated using simulated and real LADAR images. The results show the robustness of the proposed descriptor in the presence of noise, low resolution, view change, rotation, translation, and scaling effects. The results also show the high discrimination capability for the new descriptor over the traditional techniques such as invariant moments descriptor, which is used to benchmark the simulation and the experimental results.