Search

Block matching algorithms (BMAs) like sum of absolute difference (SAD) and normalised cross correlation (NCC) form the basic building block in many computer vision applications. Full search (FS) BMAs take into account every pixel in a block, and are therefore very effective for wide motion search using different blocks. However, FS is computationally very expensive and unsuitable for real-time applications. We present efficient strategies for extracting data parallelism in FS SAD and NCC algorithms with fine-grained optimisation techniques for fully exploiting the computational capacity of the NVIDIA's multicore graphic processing units (GPUs). We demonstrate that proposed parallel implementation achieves a speedup of around 34× for SAD and 59× for NCC on GPU GTX 280, compared to sequential implementation on dual socket quad core Xeon processor with 2.50 GHz and 16 GB DRAM. Since GPUs are quite cheap and popular, our algorithm can form the base for many real-time video systems.

The bi-criteria scheduling approach may require several different criteria to be considered simultaneously when evaluating the quality of solution or a schedule. In general, scheduling directed acyclic graph (DAG) in grid occupies large number of computing resources or processors. To address this problem, we present an effective bi-criteria scheduling heuristic called 'schedule optimisation with duplication-based bi-criteria scheduling algorithm' (SODA). The proposed approach comprises two stages: 1) duplication-based scheduling – optimises the primary criterion, i.e., execution time; 2) compaction of schedules – minimises the processor requirements and optimises secondary criterion, i.e., economic cost without increasing the makespan obtained in primary scheduling. The experimental results reveal that the proposed approach generates schedules with low processor requirements which are fairly optimised for both economic cost and makespan for executing DAG applications in the grid environments.

This paper presents a low-cost framework for combining multimodal information (visible, IR and audio signal) for small area surveillance and security applications. The system uses audio and video information to capture different aspects of the environment and infrared imagery is used for low lighting conditions. The visual processing module of the system uses a motion-based approach for detecting objects, and Kalman filter for tracking. Environmental sound is recognised by extracting Mel-Frequency Cepstral Coefficients (MLCCs) audio features and then classified by Dynamic Time Warping (DTW) technique. Experimental results on some typical sequences show promising results.