Applying adaptive deep learning model for detecting traffic vehicles

Abstract

Image analysis for detecting traffic vehicles is a problem in the field of computer vision. This problem has many useful applications in self-driving vehicle systems, traffic management, and vehicle flow measurement at important locations and routes. There are many approaches to solve this problem, such as contour representation, feature extraction, machine learning, and deep learning networks. In this paper, the author proposes a solution using an adaptive learning model based on a deep learning network to solve the problem. To evaluate the effectiveness of the solution, the author built a testing system based on the Darknet-53 deep learning network. The system was tested on both standard and self-collected datasets. The results show that the system achieves high accuracy and feasibility when applied to real-world applications.