OPEN ACCESS
In this paper, first, we present the problem of Non Cooperative Target Recognition (NCTR) as a supervised classification problem. After a presentation on the radar acquisition system of range profiles and the problem of recognition, followed by a statistical study of data, we use a classical classification method of K Nearest Neighbors (KNN) to do this classification. We explore and compare the performances of this algorithm based on the choice of the distances, the choice of K and the nature of used data (denoised or not). KNN algorithm has been executed initially on CPU with Matlab and then on GPU. Arithmetic operations and memory access pattern has been studied to get the best parallelization. Finally, we conclude with a discussion about possible perspectives for the proposed method especially by focusing on other representation spaces or other classification methods.
RÉSUMÉ
Dans cet article, nous nous sommes intéressés aux problèmes de reconnaissance noncoopérative de cibles (NCTR) en tant que problème de classification supervisée. Après une présentation du système d’acquisition des profils distance radar et du problème de reconnaissance, suivie d’une étude statistique des données, nous proposons d’utiliser un algorithme des K plus proches voisins (KPPV) dont les performances sont détaillées en fonction du nombre de voisins K, du type de distance utilisée et de la nature des données utilisées (débruitées ou non). Dans un second temps, cet algorithme a été parallélisé sur un processeur many-cœurs (GPU : Graphics Processing Unit). Les opérations arithmétiques et le modèle d’accès mémoire ont été étudiés pour obtenir la meilleure parallélisation des calculs. Enfin, nous terminons par une discussion autour des perspectives envisageables pour la méthode proposée, notamment en s’intéressant à d’autres espaces de représentation ou à d’autres méthodes de classification.
NCTR, KNN, GPU, HRR, radar, classification
MOTS-CLÉS
NCTR, KPPV, GPU, HRD, radar, classification
Balasubramanian M., Schwartz E. L. (2002). The Isomap Algorithm and Topological Stability. Science, vol. 295, no 5552, p. 7.
Brousseau C. (2010). Development of a tree structured hierarchical wavelet representation of synthetic database to NCTR. In Proc. IEEE Radar Conference, p. 368–373.
Chen L., Tang H. (2004). Improved computation of beliefs based on confusion matrix for combining multiple classifiers. Electronics Letters, vol. 40, no 4, p. 238 - 239.
Collange S. (2010). Analyse de l’architecture GPU Tesla. Rapport technique. DALI, ELIAUS, Université de Perpignan.
Cooke S. J., Vlasov A. N., Levush B., Chernyavskiy I. A., Antonsen T. M. (2011). GPUaccelerated 3D time-domain simulation of vacuum electron devices. In Proc. IEEE International Vacuum Electronics Conference, p. 305–306.
Duda R. O., Hart P. E., Stork D. G. (2001). Pattern Classification (2e éd.). Wiley-Interscience.
Garcia V., Debeuve E., Nielsen F., Barlaud M. (2010). K-nearest neighbor search : Fast GPUbased implementations and application to high-dimensional feature matching. In Proc. ICIP Conference, p. 3757-3760.
Jain A. K., Duin R. P. W., Mao J. (2000). Statistical pattern recognition: a review. IEEE Trans. Pattern Anal. Mach. Intell., vol. 22, no 1, p. 4–37.
Kantardzic M. (2011). Data Mining (Concepts, Models, Methods, and Algorithms). WileyInterscience.
Kirk D., Hwu W. (2010). Programming Massively Parallel Processors. NVIDIA, Elsevier Inc.
Kittler J., Society I. C., Hatef M., Duin R. P. W., Matas J. (1998). On combining classifiers. IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 20, p. 226–239.
Kotsiantis S. B. (2007). Supervised Machine Learning: A Review of Classification Techniques. Informatica, vol. 31, p. 249–268.
Lin J., Lin J. (2010). Accelerating BP Neural Network-Based Image Compression by CPU and GPU Cooperation. In Proc. ICMT Conference, p. 1–4.
Liu H., Yang Z., He K., Bao Z. (2005). Radar high range resolution profiles recognition based on wavelet packet and subband fusion. In Proc. ICASSP conference, vol. 5, p. 445-448.
Ma S., Ji C. (1999). Performance and efficiency: recent advances in supervised learning. Proc. of the IEEE, vol. 87, no 9, p. 1519–1535.
Martin A., Osswald C. (2007). Une nouvelle règle de combinaison répartissant le conflit - Applications en imagerie Sonar et classification de cibles Radar. Traitement du signal.
Moruzzis M., Colin N. (1998). Radar target recognition by Fuzzy Logic. IEEE Aerospace and Electronic Systems Magazine, vol. 13, no 7, p. 13–20.
Nimier V., Bastiere A., Colin N., Moruzzis M. (2000). MILORD, an application of multifeature fusion for radar NCTR. In Proc. 3rd Information Fusion Conference.
NVIDIA. (2007). White Paper: Accelerating MATLAB with CUDA Using MEX Files.
NVIDIA Programming Guide. (2011).
Owens J., Luebke D., Govindaraju N., Harris M., Kruger J., Lefohn A. (2007). A Survey of General-Purpose Computation on Graphics Hardware. Computer Graphics Forum, vol. 26, p. 80–113.
Rihaczek A., Herschowitz S. (2000). Theory and Practice of Radar Target Identification. Artech House.
Sawall S., Ritschl L., Knaup M., Kachelriess M. (2011). Performance comparison of OpenCL and CUDA by benchmarking an optimized perspective backprojection. In 3rd Workshop on High Performance Image Reconstruction, p. 15–8.
Skolnik M. (1962). Introduction to radar systems. McGraw-Hill.
Tait P. (2005). Introduction to Radar Target Recognition. The IEE. Tenenbaum J. B., Silva V., Langford J. C. (2000, 22). A Global Geometric Framework for Nonlinear Dimensionality Reduction. Science, vol. 290, no 5500, p. 2319–2323.
Volkov V. (2010). Better performance at lower occupancy. In GPU Technology Conference.
Wehner D. R., Barnes B. (1994). High-Resolution Radar. Artech House Publishers.
Xu F., Mueller K. (2007). Real-time 3D computed tomographic reconstruction using commodity graphics hardware. Physics in Medicine and Biology, vol. 52, no 12, p. 3405-3419.