One of the goals of planetary exploration is to cache rock samples for subsequent return to the Earth in future Mars Sample Return missions. Rocks on the Martian surface are one of the most interesting science targets for geologists and planetary scientists. Hence, it is essential to develop a method for the accurate segmenta-tion of Martian rocks in Mars images. This thesis introduces a new approach to segmenting Mars images captured by the NASA Mars Exploration Rover (MER). An improved OTSU and Canny operator are utilized for detecting rock regions and their space relations, respectively. The closed contours of detected rocks are gained by the use of template dilatation edge linking for a given set of images. These images have been obtained from MER Navcam and Pancam. Experimental results of six representative images (with different illumination levels, spectral bands and scenes) including a total of 128 rocks are shown. In these experiments qualitative and quantitative comparisons are accomplished. The results demonstrate that the proposed approach is consistent with human perception and is the best in terms of the average values over the performance indices such as Precision, Recall and misclassification error in comparison to the existing approaches. Additionally, a method is proposed for computing the size of a detected rock through the stereo triangulation technique. Experimental results also show that this proposed method offers better accuracy than the standard disparity algorithm. Currently, science target selection, and whether or not it is possible for a robot arm to touch the target, is accomplished by human operators and scientists on the Earth. The use of onboard autonomy would greatly reduce the human inter-vention, and it would be advantageous if the rover could evaluate autonomously whether the robot arm could place an instrument against an identified science target. In this thesis a fuzzy logic-based system is presented to address the problem of autonomous science target touchability evaluation. The touchability of a potential science target is assessed in terms of its size (the bounding area of the rock), SV (the science value of the target), distance (the reachable distance of the arm between its base and the science target), and orientation (the angular regions of the arm's shoulder azimuth). In particular, the plane in front of the arm is divided into a number of partitions, which are ranked with the different touchability levels by the use of a fuzzy rule-based system. Simulations on the rank of science object touchability are carried out, via hardware implementation. Based on the real data gathered from the cameras and the Schunk arm experimental results successfully verify the validity of the proposed touchability approach and associated software and hardware implementations
Autonomous Science Target Detection and Touchability Assessment for Planetary Exploration
Gui, C. (Awdur). 03 Medi 2015
Traethawd ymchwil myfyriwr: Traethawd Ymchwil Doethurol › Doethur mewn Athroniaeth