Open AccessCOMPUTER VISION BASED TRAFFIC SIGN SENSING FOR SMART TRANSPORT
Author(s) -
H. James Deva Koresh
Publication year - 2019
Publication title -
journal of innovative image processing
Language(s) - English
Resource type - Journals
ISSN - 2582-4252
DOI - 10.36548/jiip.2019.1.002
Subject(s) - convolutional neural network , computer science , reliability (semiconductor) , traffic sign recognition , traffic sign , frame (networking) , artificial intelligence , set (abstract data type) , path (computing) , real time computing , sign (mathematics) , computer vision , computer network , mathematical analysis , mathematics , power (physics) , physics , quantum mechanics , programming language
The paper puts forward a real time traffic sign sensing (detection and recognition) frame work for enhancing the vehicles capability in order to have a save driving, path planning. The proposed method utilizes the capsules neural network that outperforms the convolutional neural network by eluding the necessities for the manual effort. The capsules network provides a better resistance for the spatial variance and the high reliability in the sensing of the traffic sign compared to the convolutional network. The evaluation of the capsule network with the Indian traffic data set shows a 15% higher accuracy when compared with the CNN and the RNN.