Chenhao Wang

High-Speed Video Transfer and Real-Time Infrared Spots Detection Based on FPGA

This paper presents a high-speed video transfer scheme and a real-time infrared spots detection algorithm designed for Field Programmable Gate Array (FPGA) implementation. Rather than IEEE 1394a, two IEEE 1394b interfaces are alternatively used to ensure high-resolution image transfer in real time. In order to execute fast infrared spots detection, a parallel algorithm that processes four pixels per clock cycle is proposed. It detects infrared spots in a single pass over a frame and its implementation is only composed of combinatorial logic and registers. Furthermore, the execution time of the algorithm is independent of image content. A prototype system is implemented in an FPGA device. It is capable of transferring 1024 * 768 images smoothly at 60 fps and detecting infrared sports in a 1024 * 768 image within 1.966ms, demonstrating its superiority over the existing multi-pass algorithms and some other one-pass algorithms. Details of software and hardware architecture are discussed in this paper.

A fast approach to deformable surface 3D tracking

Deformable surface 3D tracking is a severely under-constrained problem and great efforts have been made to solve it. A recent state-of-the-art approach solves this problem by formulating it as a second order cone programming (SOCP) problem. However, one drawback of this approach is that it is time-consuming. In this paper, we propose an effective method for 3D deformable surface tracking. First, we formulate the deformable surface tracking problem as a linear programming (LP) problem. Then, we solve the LP problem with an algorithm which converges superlinearly rather than bisection algorithm whose convergence speed is linear. Our experimental studies on synthetic and real data have demonstrated the proposed method can not only reliably recover 3D structures of surfaces but also run faster than the state-of-the-art method.

Deformable Surface 3D Tracking Using Linear Programming

A method for 3D shape reconstruction of deformable surfaces from consecutive frames was presented. In our method, the model of the surface is represented by a triangulated mesh. The constraints for the model, including keypoint correspondences and disallowing large changes of edge orientation between consecutive frames, are formulated as Linear Programming (LP) constraints. Therefore the deformable surface 3D tracking method turns into an LP problem that can be effectively solved. The robustness and efficiency of our approach are validated on synthetic and real data.

FPGA IMPLEMENTATION FOR FAST INFRARED SPOTS DETECTION

This paper introduces a fast infrared spots detection algorithm designed for field-programmable gate array (FPGA) implementation. The proposed algorithm processes four pixels per clock cycle and detects infrared spots in a single pass over a frame. The implementation of the algorithm is only composed of combinatorial logic and registers. Furthermore, the execution time of the algorithm is independent of image content. For prototyping and evaluation purposes, the algorithm is implemented in an FPGA device. Demonstrated its superiority over the existing multi-pass algorithms and some other one-pass algorithms, it processes 1024u768 images smoothly at 60 fps and detects infrared sports in a 1024u768 image within 1.966ms.

Nonrigid stereo reconstruction using linear programming

Nonrigid reconstruction plays an important role in many applications like image-based modeling, human-computer interaction. In this paper, we propose an effective approach for 3D non-rigid structure reconstruction using linear programming from an image pair taken with a stereo rig. In contrast to previous approaches, the proposed method neither involves smoothness constraints nor need prior knowledge between consecutive frames, which enables us to recover shapes of surfaces with smooth, sharp and other complex deformations from a single image pair. Specifically, we model the surface as a triangulated mesh and formulate the reconstruction problem as a Linear Programming (LP) problem using L∞ that can be effectively solved. The LP problem consists of data constraints which are 3D-to-2D keypoint correspondences and shape constraints which preserve original lengths of mesh edges. The robustness and accuracy of our approach are evaluated quantitatively on synthetic data and on real data.