Deep Learning-Based Speed Bump Detection Model for Intelligent Vehicle System Using Raspberry Pi
Deepak Kumar Dewangan, Satya Prakash Sahu
Abstract
Artificial intelligence in vision based approaches have proven to be effective in various phases of intelligent vehicle system (IVS). An IVS has to intelligently take many critical decisions in heterogeneous environment. Speed bump detection is one such issue in real world due to its varying appearance in dynamic scene. The major issue is the scaling appearance of such objects from far distance and often viewed as small entity. In the proposed article, deep learning and computer vision based speed bump detection model is proposed, which assist and control the driving behavior of an IVS before it reaches to speed bump. The behavior of IVS has been explored and tested by incorporating the proposed method with a real time embedded prototype and found to be more efficient and comparable with state-of-art techniques. The overall performance of the proposed model has been achieved in terms of accuracy, precision and F-Measure as 98.54%, 99.05% and 97.89% respectively in the prepared real time environment.