Optimization and Implementation of Synthetic Basis Feature Descriptor on FPGA

Feature detection, description, and matching are crucial steps for many computer vision algorithms. These steps rely on feature descriptors to match image features across sets of images. Previous work has shown that our SYnthetic BAsis (SYBA) feature descriptor can offer superior performance to other binary descriptors. This paper focused on various optimizations and hardware implementation of the newer and optimized version. The hardware implementation on a field-programmable gate array (FPGA) is a high-throughput low-latency solution which is critical for applications such as high-speed object detection and tracking, stereo vision, visual odometry, structure from motion, and optical flow. We compared our solution to other hardware designs of binary descriptors. We demonstrated that our implementation of SYBA as a feature descriptor in hardware offered superior image feature matching performance and used fewer resources than most binary feature descriptor implementations.

[1]  Ronen Basri,et al.  Viewpoint-aware object detection and continuous pose estimation , 2012, Image Vis. Comput..

[2]  Chen-Chien James Hsu,et al.  FPGA-Based Hardware Design for Scale-Invariant Feature Transform , 2018, IEEE Access.

[3]  Javier Civera,et al.  1‐Point RANSAC for extended Kalman filtering: Application to real‐time structure from motion and visual odometry , 2010, J. Field Robotics.

[4]  Xiang Zhou,et al.  A New FPGA Architecture of FAST and BRIEF Algorithm for On-Board Corner Detection and Matching , 2018, Sensors.

[5]  George A. Constantinides,et al.  A Parallel Hardware Architecture for Scale and Rotation Invariant Feature Detection , 2008, IEEE Transactions on Circuits and Systems for Video Technology.

[6]  Ning Wu,et al.  Fast Facial Image Super-Resolution via Local Linear Transformations for Resource-Limited Applications , 2011, IEEE Transactions on Circuits and Systems for Video Technology.

[7]  Adrian G. Bors,et al.  Smoothing of optical flow using robustified diffusion kernels , 2010, Image Vis. Comput..

[8]  René Cumplido,et al.  Improving the construction of ORB through FPGA-based acceleration , 2017, Machine Vision and Applications.

[9]  Jan Faigl,et al.  Real-Time FPGA-Based Detection of Speeded-Up Robust Features Using Separable Convolution , 2018, IEEE Transactions on Industrial Informatics.

[10]  Dah-Jye Lee,et al.  Visual Odometry Drift Reduction Using SYBA Descriptor and Feature Transformation , 2016, IEEE Transactions on Intelligent Transportation Systems.

[11]  Dah-Jye Lee,et al.  Hardware Friendly Robust Synthetic Basis Feature Descriptor , 2019 .

[12]  Adam Schmidt,et al.  A High-Performance FPGA-Based Image Feature Detector and Matcher Based on the FAST and BRIEF Algorithms , 2015 .

[13]  David G. Lowe,et al.  Distinctive Image Features from Scale-Invariant Keypoints , 2004, International Journal of Computer Vision.

[14]  Rita Cucchiara,et al.  Real-time object detection and localization with SIFT-based clustering , 2012, Image Vis. Comput..

[15]  Jan-Michael Frahm,et al.  Detailed Real-Time Urban 3D Reconstruction from Video , 2007, International Journal of Computer Vision.

[16]  John N. Lygouras,et al.  Fully pipelined FPGA-based architecture for real-time SIFT extraction , 2016, Microprocess. Microsystems.

[17]  Luka Daoud,et al.  A fully pipelined FPGA accelerator for scale invariant feature transform keypoint descriptor matching , 2020, Microprocess. Microsystems.

[18]  George C. Runger,et al.  An Automated Feature Selection Method for Visual Inspection Systems , 2006, IEEE Transactions on Automation Science and Engineering.

[19]  Yung-Chang Chen,et al.  High-Performance SIFT Hardware Accelerator for Real-Time Image Feature Extraction , 2012, IEEE Transactions on Circuits and Systems for Video Technology.

[20]  Dah-Jye Lee,et al.  An efficient feature descriptor based on synthetic basis functions and uniqueness matching strategy , 2016, Comput. Vis. Image Underst..