Abstract: During the paddy hulling process, instability in operational performance frequently results in issues such as incomplete husk removal and brown rice breakage, negatively impacting both the quality of paddy processing and economic efficiency. Traditional methods for detecting the paddy hulling rate and broken brown rice rate predominantly rely on manual labor, suffering from disadvanges such as long time consumption, labor-intensive, and prone to subjective interference. The application of machine vision technology can realize rapid non-destructive detection of these parameters, playing a vital role in enhancing the quality of paddy processing. This paper reviews the advancements of static and dynamic image acquisition technologies for paddy, and discusses the applications of traditional algorithms and deep learning algorithms in image enhancement. Meanwhile, it examines the current state of research regarding the use of traditional machine learning and deep learning algorithms in the detection of paddy hulling rate and broken brown rice rate. Based on identified limitations in existing research, the future development directions of this field are prospected.