Triangle recognizing method based on maximum interior angle hash function

Abstract

The invention relates to a triangle recognizing method based on a maximum interior angle hash function. The triangle recognizing method comprises the following steps: 1, constructing a guide star catalog by adopting a star magnitude threshold value method according to a basic star catalog; constructing a triangle by using any three stars, constructing a triangle feature library by taking the maximum interior angle as a basic feature value, carrying out blocking storage according to the magnitude of interior angles, and constructing a hash function based on the maximum interior angle; 2, reading the guide star catalog and a guide feature library; preferentially recognizing brighter fixed stars during recognition, carrying out descending sorting on a star extraction result according to the magnitude of a gray value, constructing observation triangles, computing a related star angular distance, determining the maximum interior angles of the observation triangles, locating a subblock of the feature library by adopting a mode of the hash function, and carrying out the recognition on each observation triangle in the subblock, thus increasing the recognition speed. According to the triangle recognizing method disclosed by the invention, the number of feature triangles is greatly reduced; compared with a traditional triangle recognizing algorithm based on the star angular distance, the comparison times of the star angular distance are reduced, and the recognition speed is increased.

Description

technical field [0001] The invention relates to an all-celestial autonomous star map identification method, in particular to a triangle identification method based on a maximum internal angle hash function, which belongs to the field of astronomical navigation. Background technique [0002] The star sensor is currently known as the highest precision carrier attitude sensor. The star sensor solves the attitude of the vehicle by measuring the orientation of the star. Star map recognition is the key technology of star sensor, and the recognition speed, recognition rate and robustness are the key indicators to measure the performance of the recognition algorithm. The triangle algorithm based on the star angle distance is the most mature recognition algorithm in engineering, and it uses the star angle distance between two pairs as the feature quantity for matching. The characteristic dimension of the triangle algorithm is low, each observation triangle needs at least three comp...

AI Technical Summary

Problems solved by technology

The characteristic dimension of the triangle algorithm is low, each observation triangle needs at least three comparisons, and the search time is long

The triangle identification algorithm based on the main star uses the two brightest stars in the star map as the main star pair, and the other stars as auxiliary navigation stars, which can greatly reduce the number of navigation triangles and shorten the search time. The disadvantage of this type of algorithm is that when the relative stars When the magnitude is small, the recognition rate will be severely reduced, especially when the star sensor works in the atmosphere, due to weather and other reasons, the main star in the field of view (Field of View, FOV) may be missing, or the brightness of the main star is lower than other observed stars , resulting in a reduction in the recognition rate or even failure to recognize the entire star map

In view of the deficiency of the main star algorithm, Domenico proposed an improved triangle algorithm that does not rely on brightness. It is based on the large star library. By matching the angular distance of multiple triangles, the accuracy is greatly improved compared with the ordinary triangle algorithm, ensuring the accuracy of the recognition. Reliability, the disadvantage of this algorithm is that the recognition speed is slow

In addition, the triangle algorithm combined with other geometric features with translation and rotation invariant characteristics, such as inertial feature ratio, triangle inscribed circle radius, etc., has achieved good results in recognition rate and recognition speed, but in order to reduce the navigation feature library When generating the feature library, these algorithms have eliminated some navigation stars in the dense sky area, thus losing the integrity. When the field of view of the star sensor is small, the problem will be highlighted

Images