A Method of Endmember Extraction in Hyperspectral Remote Sensing Images Based on Discrete Particle Swarm Optimization (D-PSO)
ZHANG Bing1, SUN Xu1，3*, GAO Lian-ru1, YANG Li-na2，3
1. Center for Earth Observation and Digital Earth, Chinese Academy of Sciences，Beijing 100094, China 2. Institute of Remote Sensing Applications, Chinese Academy of Sciences, Beijing 100101, China 3. Graduate University of Chinese Academy of Sciences, Beijing 100049, China
Abstract：For the inaccuracy of endmember extraction caused by abnormal noises of data during the mixed pixel decomposition process, particle swarm optimization (PSO), a swarm intelligence algorithm was introduced and improved in the present paper. By re-defining the position and velocity representation and data updating strategies, the algorithm of discrete particle swarm optimization (D-PSO) was proposed, which made it possible to search resolutions in discrete space and ultimately resolve combinatorial optimization problems. In addition, by defining objective functions and feasible solution spaces, endmember extraction was converted to combinatorial optimization problem, which can be resolved by D-PSO. After giving the detailed flow of applying D-PSO to endmember extraction and experiments based on simulative data and real data, it has been verified the algorithm’s flexibility to handle data with abnormal noise and the reliability of endmember extraction were verified. Furthermore, the influence of different parameters on the algorithm’s performances was analyzed thoroughly.
[1] Keshava N, Mustard J F. IEEE Signal Processing Magazine, 2002, 19(1): 44. [2] Boardman J W, Kruse F A, Green R O. Mapping Target Signatures <i>via</i> Partial Unmixing of AVIRIS Data: Citeseer, 1995. [3] Winter M E. N-FINDR: an Algorithm for Fast Autonomous Spectral End-Member Determination in Hyperspectral Data, 1999. 266. [4] Nascimento J M, Dias J M. IEEE Transactions on Geoscience and Remote Sensing, 2005, 43(4): 898. [5] Gruninger J H, Ratkowski A J, Hoke M L. The Sequential Maximum Angle Convex Cone (SMACC) Endmember Model. Orlando, FL, USA: SPIE, 2004. 1. [6] Neville R. Automatic Endmember Extraction from Hyperspectral Data for Mineral Exploration, 1999. [7] Craig M D. IEEE Transactions on Geoscience and Remate Sensing, 1994, 32(3): 542. [8] Plaza A, Martinez P, Perez R, et al. IEEE Transactions on Geoscience and Remote Sensing, 2002, 40(9): 2025. [9] Filippi A M, Archibald R. IEEE Transactions on Geoscience and Remote Sensing, 2009, 47(3): 771. [10] Boardman J W. Geometric Mixture Analysis of Imaging Spectrometry Data. Pasadena, CA, USA, 1994. 2369. [11] Dai Q, Liu J. Application of Ant Colony Optimization (ACO) Algorithm to Remote Sensing Image Classification. Wuhan, China: SPIE, 2007. 67881A. [12] Le H M, Kallel A, Descombes X. Application of Ant Colony Optimization to Image Classification Using a Markov Model with Non-Stationary Neighborhoods. Brugge, Belgium: SPIE, 2005. 59810C. [13] Lu J. A Self-Adaptive Ant Colony Optimization Approach for Image Segmentation: SPIE, 2005. 59853F. [14] Omran M, Engelbrecht A P, Salman A. International Journal of Computational Intelligence, 2005, 2(2): 124. [15] Engelbrecht A P. Fundamentals of Computational Swarm Intelligence: John Wiley & Sons, 2006. [16] Eberhart R, Kennedy J. A New Optimizer Using Particle Swarm Theory. Nagoya, Japan, 1995. 39. [17] Heinz D C, Chang C I. IEEE Transactions on Geoscience and Remote Sensing, 2001, 39(3): 529. [18] Green A A, Berman M, Switzer P, et al. IEEE Transactions on Geoscience and Remote Sensing, 1988, 26: 65. [19] Chang C, DU Q. IEEE Transactions on Geoscience and Remote Sensing, 2004, 42(3): 608. [20] Swayze G, Clark R, Kruse F, et al. Ground-Truthing Aviris Mineral Mapping at Cuprite, Nevada. United States, 1992. 47.
