Anisotropy Diffusion Kuwahara filtering and Dual-discriminator D2C Conditional Generative Adversarial Network Classification on Spatio-Temporal Transportation’s Traffic images
Tukaram K. Gawali, Shailesh Shivaji Deore · 2024
The Spatio-temporal transportations have various issues like traffic congestions, weather issues and wind directions. The major problems are to prevent from traffic-based accidents. The traffic may be in homogenous and heterogeneous format. In this paper the complete focus is based on heterogeneous traffic flow. As per the huge demands of vehicles either are petrol, diesel or electrical, the densities of vehicles are increases rapidly along with the population. India is the largest populated country in the world and having more vehicles as comparative roads availability. The whole world always faces issues to identify traffic condition with images. In our proposed system, the source image is extracted from a dataset of road vehicle images. To enhance the image quality, noise reduction is applied using Anisotropy Diffusion Kuwahara filtering during the pre-processing stage. The classification using DCGAN-TDR is implemented to generate 97.74% accuracy with high precision.