DimensioNet: A Real-Time System for Object Detection, Dimension Estimation, and Dominant Color Identification Using YOLOv8

Abstract: This research presents DimensioNet, a real-time object detection and analysis system capable of identifying objects, estimating their physical dimensions, and extracting dominant color information from live video feeds. Developed using the YOLOv8 deep learning framework, the system integrates advanced computer vision techniques to provide accurate, efficient, and interpretable results. Object dimensions are computed using bounding box analysis and a calibrated scaling factor, while dominant color is determined by analyzing the average pixel values of detected objects. The system employs Python libraries such as OpenCV and Matplotlib to enable real-time frame processing and visual annotation. DimensioNet is designed for practical deployment in sectors like logistics, manufacturing, and e-commerce, where rapid and automated object analysis is essential. Evaluation of the system demonstrates high precision in detection, reliable color identification, and consistent dimension estimation. The work lays a foundation for future improvements, including multi-object tracking, integration with IoT devices, and adaptive real-world applications.