Research on Novel Lightweight Object Detection and Recognition Algorithms
Qi He, Hao Shen, Yang Jin · 2023
Object detection have broad applications in industrial production, medicine, and people’s lives. It is also an important research direction in the field of computer vision. Exploring the application of target detection technology has important practical significance. This paper focuses on detecting safety helmet-wearing in the construction site scene and manufacturing. Our paper presents novel and lightweight algorithms for detecting helmet-wearing. Our work has three main contributions. Firstly, we adopt MobileNet as a backbone feature extraction network with fewer parameters. Secondly, our algorithm uses DSC to reduce the number of parameters. Moreover, we add an AM to our methods. The methods we proposed have been proven to be effective through experiments.