Exploration architecturale pour la conception d'un système sur puce de vision robotique, adéquation algorithme-architecture d'un système embarqué temps-réel

La problematique de cette these se tient a l'interface des domaines scientifiques de l'adequation algorithme architecture, des systemes de vision bio-inspiree en robotique mobile et du traitement d'images.Le but est de rendre un robot autonome dans son processus de perception visuelle, en integrant au sein du robot cette tâche cognitive habituellement deportee sur un serveur de calcul distant.Pour atteindre cet objectif, l'approche de conception employee suit un processus d'adequation algorithme architecture, ou les differentes etapes de traitement d'images sont analysees minutieusement.Les traitements d'image sont modifies et deployes sur une architecture embarquee de facon a respecter des contraintes d'execution temps-reel imposees par le contexte robotique.La robotique mobile est un sujet de recherche academique qui s'appuie sur des approches bio-mimetiques.La vision artificielle etudiee dans notre contexte emploie une approche bio-inspiree multi-resolution, basee sur l'extraction et la mise en forme de zones caracteristiques de l'image.Du fait de la complexite de ces traitements et des nombreuses contraintes liees a l'autonomie du robot, le deploiement de ce systeme de vision necessite une demarche rigoureuse et complete d'exploration architecturale logicielle et materielle.Ce processus d'exploration de l'espace de conception est presente dans cette these.Les resultats de cette exploration ont mene a la conception d'une architecture principalement composee d'accelerateurs materiels de traitements (IP) parametrables et modulaires, qui sera deployee sur un circuit reconfigurable de type FPGA.Ces IP et le fonctionnement interne de chacun d'entre eux sont decrits dans le document.L'impact des parametres architecturaux sur l'utilisation des ressources materielles est etudie pour les traitements principaux.Le deploiement de la partie logicielle restante est presente pour plusieurs plate-formes FPGA potentielles.Les performances obtenues pour cette solution architecturale sont enfin presentees.Ces resultats nous permettent aujourd'hui de conclure que la solution proposee permet d'embarquer le systeme de vision dans des robots mobiles en respectant les contraintes temps-reel imposees.

[1]  Anton Kummert,et al.  Implementation of hardware-optimized 3-D wave digital filters for motion-based object detection in video scenes , 2008, 2008 16th European Signal Processing Conference.

[2]  John F. Canny,et al.  A Computational Approach to Edge Detection , 1986, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  François Verdier,et al.  A Modular SystemC RTOS Model for Embedded Services Exploration , 2007 .

[4]  A. Dupret,et al.  On chip vision system architecture using a CMOS retina , 2004, IEEE Intelligent Vehicles Symposium, 2004.

[5]  Lester J. Kozlowski.  Noise minimization via deep submicron system-on-chip integration in megapixel CMOS imaging sensors , 2005, Optics + Optoelectronics.

[6]  Christopher G. Harris,et al.  A Combined Corner and Edge Detector , 1988, Alvey Vision Conference.

[7]  Alexandre Bernardino,et al.  Vergence control for robotic heads using log-polar images , 1996, Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems. IROS '96.

[8]  Yamin Li,et al.  Cost/performance tradeoff of n-select square root implementations , 2000, Proceedings 5th Australasian Computer Architecture Conference. ACAC 2000 (Cat. No.PR00512).

[9]  M. Ishikawa,et al.  Digital vision chips and high-speed vision systems , 2001, 2001 Symposium on VLSI Circuits. Digest of Technical Papers (IEEE Cat. No.01CH37185).

[10]  G LoweDavid,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004 .

[11]  Cordelia Schmid,et al.  Scale & Affine Invariant Interest Point Detectors , 2004, International Journal of Computer Vision.

[12]  Cordelia Schmid,et al.  Local Grayvalue Invariants for Image Retrieval , 1997, IEEE Trans. Pattern Anal. Mach. Intell..

[13]  S. Sarkar,et al.  A bio-inspired model for multi-scale representation of even order Gaussian derivatives , 2004, Proceedings of the 2004 Intelligent Sensors, Sensor Networks and Information Processing Conference, 2004..

[14]  Matthew French.  A Power Efficient Image Convolution Engine for Field Programmable Gate Arrays , 2004 .

[15]  Miriam Leeser,et al.  Smart camera based on reconfigurable hardware enables diverse real-time applications , 2004, 12th Annual IEEE Symposium on Field-Programmable Custom Computing Machines.

[16]  Gary Smith,et al.  High-Level Synthesis: Past, Present, and Future , 2009, IEEE Design & Test of Computers.

[17]  Giulio Sandini,et al.  On the Advantages of Polar and Log-Polar Mapping for Direct Estimation of Time-To-Impact from Optical Flow , 1993, IEEE Trans. Pattern Anal. Mach. Intell..

[18]  Dominique Ginhac,et al.  A high speed programmable focal-plane SIMD vision chip , 2010 .

[19]  Jocelyn Sérot,et al.  A New High-Level Methodology for Programming FPGA-Based Smart Camera , 2010, 2010 13th Euromicro Conference on Digital System Design: Architectures, Methods and Tools.

[20]  J. Crowley,et al.  Fast Computation of Characteristic Scale Using a Half-Octave Pyramid , 2002 .

[21]  Ángel Rodríguez-Vázquez,et al.  ACE16k: the third generation of mixed-signal SIMD-CNN ACE chips toward VSoCs , 2004, IEEE Transactions on Circuits and Systems I: Regular Papers.

[22]  Bernd Girod,et al.  CHoG: Compressed histogram of gradients A low bit-rate feature descriptor , 2009, 2009 IEEE Conference on Computer Vision and Pattern Recognition.

[23]  Nick G. Kingsbury,et al.  Multiscale Keypoint Detection using the Dual-Tree Complex Wavelet Transform , 2006, 2006 International Conference on Image Processing.

[24]  Martin D. Levine,et al.  A Review of Biologically Motivated Space-Variant Data Reduction Models for Robotic Vision , 1998, Comput. Vis. Image Underst..

[25]  Chu Kiong Loo,et al.  FPGA Implementation of Log-polar Mapping , 2008, 2008 15th International Conference on Mechatronics and Machine Vision in Practice.

[26]  Yun Zhang,et al.  Hierarchical Classification for Imbalanced Multiple Classes in Machine Vision Inspection , 2007, Fourth International Conference on Image and Graphics (ICIG 2007).

[27]  Filiberto Pla,et al.  Designing the Lattice for Log-Polar Images , 2003, DGCI.

[28]  Wen Gao,et al.  Illumination normalization for robust face recognition against varying lighting conditions , 2003, 2003 IEEE International SOI Conference. Proceedings (Cat. No.03CH37443).

[29]  Mitsumasa Koyanagi,et al.  Smart Vision Chip Fabricated Using Three Dimensional Integration Technology , 2000, NIPS.

[30]  Philippe Gaussier,et al.  Perception as a Dynamical Sensori-Motor Attraction Basin , 2005, ECAL.

[31]  Jun Yu,et al.  An Embedded Vehicular Controller with ARM and DSP for a Vision-Based AGV , 2008, 2008 International Conference on Embedded Software and Systems Symposia.

[32]  François Berry,et al.  Harnessing a multi-sensor fpga-based smart camera: a virtual processor-based approach , 2009 .

[33]  Sonia Khatchadourian,et al.  Mise en oeuvre d'une architecture de reconnaissance de formes pour la détection de particules à partir d'images atmosphériques , 2010 .

[34]  Eduardo Ros Vidal,et al.  Visual Processing Platform Based on Artificial Retinas , 2007, IWANN.

[35]  François Berry,et al.  Embedded Active Vision System Based on an FPGA Architecture , 2007, EURASIP J. Embed. Syst..

[36]  Hans P. Moravec.  Obstacle avoidance and navigation in the real world by a seeing robot rover , 1980 .

[37]  Niko Sünderhauf,et al.  COMPARING SEVERAL IMPLEMENTATIONS OF TWO RECENTLY PUBLISHED FEATURE DETECTORS , 2007 .

[38]  Masatoshi Ishikawa,et al.  Vision chip architecture using general-purpose processing elements for 1 ms vision system , 1997, Proceedings Fourth IEEE International Workshop on Computer Architecture for Machine Perception. CAMP'97.

[39]  Mariano Fons,et al.  Run-time self-reconfigurable 2D convolver for adaptive image processing , 2011, Microelectron. J..

[40]  James Archibald,et al.  FPGA implementation of vision algorithms for small autonomous robots , 2005, SPIE Optics East.

[41]  Bernhard Rinner,et al.  The evolution from single to pervasive smart cameras , 2008, 2008 Second ACM/IEEE International Conference on Distributed Smart Cameras.

[42]  David G. Lowe,et al.  Object recognition from local scale-invariant features , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[43]  Filiberto Pla,et al.  Log-polar mapping template design: From task-level requirements to geometry parameters , 2008, Image Vis. Comput..

[44]  Dah-Jye Lee,et al.  A Fast and Accurate Tensor-based Optical Flow Algorithm Implemented in FPGA , 2007, 2007 IEEE Workshop on Applications of Computer Vision (WACV '07).

[45]  E. L. Schwartz,et al.  Spatial mapping in the primate sensory projection: Analytic structure and relevance to perception , 1977, Biological Cybernetics.

[46]  Antonio Martínez-Álvarez,et al.  Automatic Generation of Bio-inspired Retina-Like Processing Hardware , 2005, IWANN.