System and method of crater detection and registration using marked point processes, multiple birth and death methods and region-based analysis
The present invention relates to a novel method and system for crater detection in planetary data based on marked point processes (MPP), effective for various object detection tasks in Earth observation, and for planetary image registration. The resulting spatial features are exploited for registration, together with fitness functions based on the MPP energy, on the mean directed Hausdorff distance, and on the mutual information. Two different methods—one based on birth-death processes and region-of-interest analysis, and the other based on graph cuts and decimated wavelets—are included within the present framework. Experimental results confirmed the effectiveness of the present invention in terms of crater detection performance and sub-pixel registration accuracy.
1. A system to detect craters on a planetary surface, the system comprising:
a sensor which captures images on the planetary surface, as input images;
a pre-processing module which crops said input images and reference images, and reduces noise and extracts contours from said input images and said reference images;
a crater extraction module which reduces and transforms image data from said preprocessing module, and performs energy minimization on said input images and said reference images using marked point processes, combined with one of a multiple-birth-and-death or multiple-birth-and-cut method;
wherein a crater-based image registration module performs a first registration by one of a crater-to-contour registration or a crater-to-crater registration;
wherein said crater-to-contour registration includes matching the craters extracted from said reference images with edge contours extracted from said input images; and
wherein crater-to-crater registration includes matching the craters extracted from both said reference images and said input images; and
a mutual information-based image registration module which performs a second registration by maximizing mutual information between said reference images and said input images to achieve a final transformation and registered images.
2. The system of claim 1 , wherein said pre-processing module further comprises:
an edge detector which extracts said contours from said input images and said reference images.
3. The system of claim 2 , wherein said pre-processing module further comprises:
a noise filter which reduces said noise by applying a Gaussian filtering to said input images and said reference images.
4. The system of claim 1 , the system further comprising:
a spacecraft including:
a storage which stores said image data on said input images; and
an onboard processing system which retrieves said input images from said storage and transfers them to a ground segment.
5. The system of claim 4 , wherein said ground segment comprises:
an antenna which receives said image data from said onboard processing system and forwards said image data to said pre-processing module;
a plurality of image registration modules including a crater detection module, said crater-based image registration module, and said mutual information-based image registration module; and
at least one storage which stores said input images, said reference images, and said registered images.
6. The system of claim 1 , wherein said sensor includes an imaging system, the imaging system being one of a thermal emission imaging system or a camera.