International Journal Of Imaging

Abstract

Vechile detection and speed detection

Author: G.Madhulika, J.Chandhana, D.Sai kiran, K.Sukumar, Ms.Navya Sree,

Abstract:

Recent explosions in the processing power of commonplace computers have allowed widespread use of deep learning techniques for traffic data analysis. security camera footage. Basic elements of traffic analysis include anomaly detection, traffic flow forecasting, vehicle re-identification, and vehicle tracking. Predicting traffic flows, sometimes known as estimating vehicle speeds, is one of the most actively studied applications in the field today. The ability to more accurately predict the need for public transportation will greatly benefit road design and reduce the likelihood of accidents. Our solution for the 2018 NVIDIA AI City Challenge aims to offer an effective method for predicting vehicle speed by combining state-of-the-art deep learning models with traditional computer vision methodologies. In this article, we provide our solution for Track 1 of the Challenge and several state-of-the-art methods for estimating vehicle speed, detecting vehicles, and tracking objects.