Skip to Main Content
Computer vision applications in the real time perspective has to meet several performance requirements but can be restrained by computing resources. To attain real time performance particularly on mobile platforms we need to apply optimized algorithms. This paper presents work on performance optimization of general computer vision algorithms such as Viola Jones Face Detection on embedded systems with limited resources. The Viola Jones algorithm which is popular for face detection, can be implemented on mobile platforms. The algorithms are benchmarked on the Intel processor and BeagleBoard xM, which is a new low-cost low-power platform based on the Texas Instruments (TI) DM 3730 processor architecture. The DM 3730 processor is characterized by the presence of an asymmetric dual-core architecture, which including an ARM and a DSP along with a shared memory between them. OpenCV, which is a famous open source computer vision library developed by Intel corporation was utilized for some of the algorithms. Comparative results for the different platforms are introduced and analyzed with an emphasis on real-time Application.