Object Detection Using YOLOv4 and PSO-CNN

Object detection, a crucial task in computer vision, involves identifying and locating objects within images or videos. In this paper a novel approach to detect object is proposed using Hyperbolic Tangent YOLOv4 and Particle Swarm Optimization (PSO) based Convolutional Neural Networks (CNNs) optimized. Initially, noise is removed by utilizing Non-Local Means Denoising algorithm and the Histogram Equalization algorithm is used for image enhancement to ensure optimal data quality. HT-YOLOv4 efficiently localizes objects within images, providing precise bounding box coordinates. These coordinates are then fed into the PSO-based CNN for accurate object classification. Experimental evaluations validate its effectiveness, highlighting its potential across diverse real-world applications in computer vision and beyond. This comprehensive methodology represents a significant advancement in object detection techniques, leveraging the strengths of cutting-edge methodologies while ensuring robustness and precision.