Researchers propose a rear-view camera system with AI object detection to augment rear-view image information, with an “image processing chip with wide-angle image distortion correction and an image buffer controller. image, a low-power KL520 chip, and a MobileNetV2-YOLOV3-Optimized (MNYLO) optimized AI model.
“According to a study by the Insurance Institute for Highway Safety (IIHS), the rate of driving collisions using only the rear view camera system is lower than using both the rear view camera system and radar. In this article, we have implemented a rear view camera system with artificial intelligence object detection to increase the information of the rear view image.Our system consists of an image processing chip (IPC ) with wide-angle image distortion correction and a frame buffer controller, a low-power KL520 chip, and a MobileNetV2-YOLOV3-Optimized (MNYLO) artificial intelligence model. experience show the three advantages of our system.Firstly,through IPC image distortion correction,we can restore the distorted inverted image.Secondly,using public data set and images collected from various times in For training artificial intelligence models, our system does not need to use image algorithms that eliminate bad weather such as rain, fog and snow to restore polluted images. Objects can still be detected by our system in weather-contaminated images. Third, compared with the AI model Tiny_YOLOV3, not only the parameters of our MNYLO have been reduced by 72.3%, the amount of computation has been reduced by 86.4%, but the object detection rate has also been reduced. been maintained and avoided sharp declines. »
Find the technical document in free access here. Published in January 2022.
Hung, K.-C. ; Lin, M.-C.; Lin, S.-F. A new energy-saving reversing camera system with artificial intelligence object detection. Electronics 2022, 11, 282. https://doi.org/10.3390/electronics11020282.
Visit the Semiconductor Engineering Technical Article Library here and discover many more academic articles about the chip industry.