DocumentCode :
3585440
Title :
Artificial Target Location on Mars Exploration
Author :
Ying Li ; Jing Peng ; Ying Du
Author_Institution :
Inst. of Space Syst. Eng., China Acad. of Space Technol., Beijing, China
Volume :
2
fYear :
2014
Firstpage :
80
Lastpage :
83
Abstract :
A new location method is proposed to locate the artificial targets on curiosity rover. Firstly, considering the influence of dust particles and non-uniform illumination on Mars, median filtering and adaptive thresholding image segmentation based on integral image are used for edge detection. Then, a two-step location method is proposed to recognize and locate these artificial targets: least square ellipse fitting is used for preliminary location and Hough transform is used for precise location. Experiments are carried out based on artificial target images of curiosity rover, which indicate that the method proposed can be used on the targets with largely deformation, and the location accuracy is within 1 pixel. The method is robust to the dust adhesion, and illumination and shadows effect. It can satisfy the Mars exploration system requirements of resolution and robustness.
Keywords :
Mars; astronomical image processing; curve fitting; edge detection; image segmentation; least squares approximations; median filters; planetary rovers; Hough transform; Mars exploration; adaptive thresholding image segmentation; artificial target images; artificial target location; curiosity rover; dust adhesion; dust particles; edge detection; illumination; integral image; least square ellipse fitting; median filtering; nonuniform illumination; two-step location method; Filtering; Fitting; Image edge detection; Mars; Noise; Robustness; Transforms; Artificial target location; Illumination robust; Mars exploration; curiosity rover;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computational Intelligence and Design (ISCID), 2014 Seventh International Symposium on
Print_ISBN :
978-1-4799-7004-9
Type :
conf
DOI :
10.1109/ISCID.2014.220
Filename :
7081942
Link To Document :
بازگشت