TY - JOUR AU - Youngyai, Chayanin AU - Suntiamorntut, Wannarat PY - 2020 TI - The Implement of Vehicle Blind Spot Detection System Using SDSoC JF - Naresuan University Journal: Science and Technology (NUJST); Vol 28 No 3 (2020): July-September 2020 DO - 10.14456/nujst.2020.30 KW - N2 -         Embedded system and image processing is the technology used to apply for designing systems to facilitate drivers in the future by replacing wing mirrors with the camera to detect objects at the blind spot at the back. When the object is detected within the blind spot zone, the system is alarm drivers. These are a concept of The blind spot detection vehicle system (BSDS). This research presents BSDS to help drivers, which is designed on Field Programmable Gate Array (FPGA). There are two main modules for creating. The first module is pre-processing image using grayscale to resize the dataset in terms of the bit stream in every pixel and region-of-interest (ROI) for specifying the area of pixels wanted to reduce the noise of the data set or data quantity that needs to use in calculation. The second module is vehicle detection and alarm by Sobel operator for object and shadow detection to detect the shadow of the vehicle and confirm the existence of it. This research shows details of each algorithm, flow chart, and effectiveness of vehicle detection. This idea has been designed and developed by a software called Software-Defined System-On-Chip (SDSoC) on the Zybo board. In conclusion, the system can detect vehicles at resolution 1920*1080 within 12 ms/frame and accuracy of vehicle detection at 100% recall. Keywords: SDSoC, FPGA, BSDS, image-processing, pre-processing image, region-of-interest, Sobel operator, shadow detection UR - https://www.journal.nu.ac.th/NUJST/article/view/Vol-28-No-3-2020-96-109