This article narrates the traffic sign recognition by using morphological hat transform and YOLOV5. Recognition of traffic signs while driving on the road is necessary to estimate traffic situations and to avoid the accident. Due to bad weather conditions and night time, it is difficult to detect traffic signs in low contrast images. To tackle this problem, in this article, morphological hat transform used to enhance low contrast images and YOLOV5 used for detection. For this traffic sign detection, YOLOV5 network is trained with 4 classes of traffic sign dataset, which contains totally 740 images, in these 592 images used for training and 148 images used as validation images. This YOLOV5 network detected traffic signs with precision 77.9, recall 93.0 and obtained 0.78 mAP.
CITATION STYLE
Vasanthi, P., & Mohan, L. (2022). YOLOV5 and Morphological Hat Transform Based Traffic Sign Recognition. In Lecture Notes in Networks and Systems (Vol. 434, pp. 579–589). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-19-1122-4_60
Mendeley helps you to discover research relevant for your work.