Vision based terrain relative navigation sensor

Abstract

Space field has witnessed vigorous change due to evolution of image processing. For precise spacecraft navigation and landing, high end image processing algorithm has to be executed in real time. In this paper, vision based in plane relative positioning and velocity sensor is discussed for precise landing. Image acquisition of the planet from the camera at the altitude of 6km–200m at successive time interval is planned. These images are used for the estimation of velocity and relative position using phase correlation based method. Image FFT are performed to compute correlation using radiation hardened FPGA. Sub pixel-shift of 1/20th pixel between two images is calculated from correlation peak using interpolation technique. Terrain relative horizontal velocity is estimated from the images pixel-shift per unit time with the help of altimeter data. This measurement technique is robust against image illumination variation, rotation tolerance of 0.3 degree between two successive images and Jitter of the space craft. The computation time is assessed from the simulation and optimized to get a faster update rate. A proto model of the sensor is tested successfully in laboratory environment using a scene simulation of lunar terrain in front of the sensor.