Pre-Processing Of CCD Star Sensor Image Based On FPGA

Star sensor is a high-accuracy instrument for attitude determination. Using the star images shoot by CCD(Charge Coupled Device)camera, The CCD Star sensor is capable of calculating the instantaneous direction of axis of the star sensor in inertia space through such steps as star detection, identification of star images and ascertainment of position, so it can decide the three abstract axes position of the carrier. The pre-processing of CCD star images has been studied in the thesis and to be realized with FPGA(Field Programmable Gate Array). The main contents are as follows:Analyzing various factors on the detection accuracy, The improvement of the single star accuracy is very important. So it's necessary to adopt effective noise filtering technology, the best target star pick-up and centroid locating algorithm and to be realized with FPGA.Based on the CCD noise model, analyzes the producing mechanism and the features of the CCD noise. The elimination ability of CDS(Correlated Double Sample) technique to reset noise and other noise is analyzed.Using improved patulous cell signing algorithm, the target star is picked up. The simulation is carried out sufficiently to analyze various centroid locating algorithm. The simulation results show that the threshold centroid algorithm is the most appropriate method for centroid locating.The pre-processing algorithm is realized with FPGA, The six realizing Verilog modules is analyzed and explained in detail. The star image produced by simulation is eliminated noises simply, target star of the star image is picked up, and centroid of star is located, The simulation result show that algorithm and realization is very efficient.