Abstract: In response to the problems which exist in traditional triangle star identification algorithms including redundant matching and poor anti-noise performance, an efficient star identification algorithm based on shape factor feature was proposed. The algorithm presents the shape factor characteristic parameter of triangle and the direction information based on traditional triangle algorithm. Based on the traditional triangle algorithm, this algorithm introduces the feature parameters of shape factor and the direction information, and selects 4 observation stars in the field of view to form a pair of observation triangles, and then performs star point matching and recognition. Compared with the traditional triangle algorithm, the algorithm proposed increases the feature information amount of the star map when identifying star, reduces the redundancy when matching, and has the advantages of small storage space and high identification speed. The experimental results show that under the simulation conditions of standard deviation of 2 pixels and standard noise of 0.7 magnitude at the star position, the identification rate of the algorithm is more than 99%. Through the physical verification of the ground experiment, the average time of self-identification of star map is about 47 ms in the operation based on the FT-C6713 DSP hardware platform of 300 MHz, which has obvious advantages over traditional triangle star identification algorithm.