Novel guide star optimal selection algorithm for star sensors based on star clustering

2018 ◽  
Vol 32 (34n36) ◽  
pp. 1840089
Author(s):  
Feng Wu ◽  
Xifang Zhu ◽  
Ruxi Xiang ◽  
Qiuyang Yu ◽  
Tingting Huang ◽  
...  
Keyword(s):  
Mean Shift ◽  
Simple Calculation ◽  
Field Of View ◽  
Space Vehicles ◽  
Optimal Selection ◽  
Selection Algorithm ◽  
Mean Shift Algorithm ◽  
Star Catalog ◽  
Celestial Sphere ◽  
Guide Star

Modern space vehicles face the challenges to obtain more and more accurate attitudes in order to complete the demanding tasks. Onboard star sensors which identify the observed stars in the field of view according to the loaded guide star catalog and output accurate attitude have attracted most interests. Guide stars are usually required to distribute uniformly on the celestial sphere to improve the performance of the star pattern identification. An optimal selection algorithm is proposed to achieve an even distribution of guide stars in this paper. Constellation features are discussed. The mean shift algorithm is analyzed. The idea that distributes stars in the local field of view to constellations is proposed by using the star pair angular separations according to the star positions in the inertial coordinate system. The optimal selection algorithm of guide stars based on star clustering is developed. Its detailed implement procedures are introduced completely. The guide star optimal selection experiment in visible band by using SAO star catalog as the original star data is implemented. It proves that the proposed algorithm has the virtue of simple calculation and easy realization. The obtained guide star distribution is superior to the regression selection algorithm and the magnitude weighted method.

Optimization of guide star catalog for daytime star sensors

2019 ◽  
Vol 34 (01n03) ◽  
pp. 2040065
Author(s):  
Feng Wu ◽  
Xifang Zhu ◽  
Qingquan Xu ◽  
Ruxi Xiang ◽  
Qiuyang Yu ◽  
...  
Keyword(s):  
Signal To Noise Ratio ◽  
Simple Calculation ◽  
High Signal ◽  
Star Sensor ◽  
Infrared Band ◽  
Star Catalog ◽  
Celestial Sphere ◽  
Guide Star ◽  
Star Pattern Identification ◽  
Navigation Information

Daytime star sensor provides accuracy navigation information to air vehicles near the ground in the daytime by observing stars. It has been an important development of modern star sensors. In order to achieve a high signal-to-noise ratio, daytime star sensors work in the infrared band to avoid interferences from sky background. Daytime star sensors output accurate attitudes by identifying the observed stars in the field of view (FOV) according to the loaded guide star catalog. Guide stars are usually required to be distributed uniformly on the celestial sphere to improve the performance of star pattern identification. The parameters including limiting magnitude and FOV are determined by processing the 2MASS star catalog as the original star data and performing star distribution statistics. After constellation features are discussed, the idea of distributing stars in the local FOV to constellations is put forward by using the star pair angular separations. An optimization algorithm to build the guide star catalog for daytime stars is proposed to achieve evenly distributed guide stars. The guide star catalog is established and analyzed, proving that the proposed algorithm has simple calculation and easy realization. The Boltzmann entropy of obtained guide star catalog drops two orders of magnitude. Guide stars are distributed more uniformly.

A Research on Navigation Star Catalog Selection Algorithm Based on SVM

2013 ◽  
Vol 706-708 ◽  
pp. 613-617
Author(s):  
Fu Cheng Liu ◽  
Zhao Hui Liu ◽  
Wen Liu ◽  
Dong Sheng Liang ◽  
Kai Cui ◽  
...  
Keyword(s):  
Support Vector Machine ◽  
Field Of View ◽  
Support Vector ◽  
Selection Algorithm ◽  
Boltzmann Entropy ◽  
Local Criterion ◽  
Star Catalog ◽  
Uniform Sampling ◽  
Global Criterion ◽  
Small Capacity

A navigation star catalog (NSC) selection algorithm via support vector machine (SVM) is proposed in this paper. The sphere spiral method is utilized to generate the sampling boresight directions by virtue of obtaining the uniform sampling data. Then the theory of regression analysis methods is adopted to extract the NSC, and an evenly distributed and small capacity NSC is obtained. Two criterions, namely a global criterion and a local criterion, are defined as the uniformity criteria to test the performance of the NSC generated. Simulations show that, compared with MFM, magnitude weighted method (MWM) and self-organizing algorithm(S-OA), the Boltzmann entropy (B.e) of SVM selection algorithm (SVM-SA) is the minimum, to 0.00207. Simultaneously, under the conditions such as the same field of view (FOV) and elimination of the hole, both the number of guide stars (NGS) and standard deviation (std) of SVM-SA is the least, respectively 7668 and 2.17. Consequently, the SVM-SA is optimal in terms of the NGS and the uniform distribution, and has also a strong adaptability.

Visual Detection Technology of the Driver's State of Fatigue Based on Multi-Information Fusion

2013 ◽  
Vol 380-384 ◽  
pp. 3925-3929
Author(s):  
Li Ming Zhao ◽  
Yi Min Zou
Keyword(s):  
Information Fusion ◽  
Particle Filtering ◽  
Visual Detection ◽  
Mean Shift ◽  
Image Features ◽  
Selection Algorithm ◽  
Color Information ◽  
Shape Model ◽  
Mean Shift Algorithm ◽  
Detection Technology

This paper attempts to combine the image features of shape, movement and color, taking the oval as the shape model of the target to immediately catch the edge of the target, using the feature light stream technology to obtain the movement information of the target. The color information can be obtained using the color histogram features of the target and mean shift algorithm. In the tracking of the color, SDA color spatial selection algorithm is introduced to improve the target and background differentiation. Then, the above information would be integrated through particle filtering technology. The self-adapting integration technology is used, and each image features weight would be decided according to its validity, thus, a quick and robust facial tracking algorithm is gained.

scholarly journals Improvement and Comparison of Mean Shift Tracker using Convex Kernel Function and Motion Information

2013 ◽  
Vol 2 (3) ◽  
Author(s):  
Shilpa Mahesh Wakode
Keyword(s):  
Kernel Function ◽  
Moving Objects ◽  
Mean Shift ◽  
Adaptive Method ◽  
Object Motion ◽  
Target Localization ◽  
Field Of View ◽  
Motion Information ◽  
Tracking Errors ◽  
Mean Shift Algorithm

Any tracking algorithm must be able to detect interested moving objects in its field of view and then track it from frame to frame. The tracking algorithms based on mean shift are robust and efficient. But they have limitations like inaccuracy of target localization, object being tracked must not pass by another object with similar features i.e. occlusion and fast object motion. This paper proposes and compares an improved adaptive mean shift algorithm and adaptive mean shift using a convex kernel function through motion information. Experimental results show that both methods track the object without tracking errors. Adaptive method gives less computation cost and proper target localization and Mean shift using convex kernel function shows good results for the tracking challenges like partial occlusion and fast object motion faced by basic Mean shift algorithm.

Automatic vertebra segmentation based on graph cut and mean shift algorithm

2011 ◽  
Vol 31 (3) ◽  
pp. 760-762
Author(s):  
Ji LIU ◽  
Xiao-dong KANG ◽  
Fu-cang JIA
Keyword(s):  
Mean Shift ◽  
Graph Cut ◽  
Mean Shift Algorithm

The mean shift algorithm and its relation to kernel regression

2016 ◽  
Vol 348 ◽  
pp. 198-208 ◽  
Author(s):  
Youness Aliyari Ghassabeh ◽  
Frank Rudzicz
Keyword(s):  
Mean Shift ◽  
Kernel Regression ◽  
Mean Shift Algorithm ◽  
The Mean

Video Vehicle Tracking Based on Improved Mean-Shift Algorithm

2011 ◽  
Vol 179-180 ◽  
pp. 1408-1411
Author(s):  
Wei Bin Chen ◽  
Xin Zhang ◽  
Su Qin Luo
Keyword(s):  
Clustering Algorithm ◽  
Mean Shift ◽  
Vehicle Tracking ◽  
Absolute Difference ◽  
Mean Shift Algorithm ◽  
Invariant Moments ◽  
Mean Shift Clustering ◽  
Video Frames ◽  
Detection Technology ◽  
Background Area

An improved Mean-Shift-based Video vehicle tracking algorithm was proposed and which can improve the real-time and accuracy of the vehicle detection technology in the application. First, it eliminates the disturbance from unrelated background by mathematical morphology operation between a traffic image and the mask of fixed background area .Then the image sequences are simulated by absolute difference of adaptive threshold for detecting latent target. At last, clusters video frames with similar characteristics which are regarded of the invariant moments vectors by Mean Shift clustering algorithm. Experimental results shown that the improved algorithm has advantages of reducing king region of vehicle matching and vehicle complete occlusion.

Sign in / Sign up

Export Citation Format

Share Document