<p>Plaque detection in dentistry requires accurate spatial localization of the plaque in respect to other oral structures. Since they are integral and cannot distinguish between small but clinically important changes in plaque area, traditional plaque indices are challenging to utilise. The conventional methods of measuring dental plaque are not reliable, accurate, sensitive, specific, or objective. As a result, they need the most meticulous clinical trial design and depend heavily on costly and time-consuming intra- and inter-examiner labelling. This study presents a novel method for identifying the tooth plaque region in a dental image by using an adaptive thresholding-based colour reduction technique. This work tackles numerous quantitative methods for automating dental plaque recognition in digital pictures using different colour models, clustering-based segmentation techniques, and the proposed colour reduction algorithm. Accurate measurement of plaque coverage was sought by evaluating a variety of computer-aided or computer-based segmentation methodologies to dental plaque images for automation. By combining many color models and clustering-based segmentation methods, the suggested method provides reliable plaque recognition in a range of image quality and illumination scenarios. With a mean Intersection over Union (IoU) of 87.4% and a sensitivity of 91.2%, the evaluation on a dataset of annotated dental pictures showed significant gains in segmentation accuracy, surpassing traditional k-means and Otsu thresholding techniques by 12.6% and 9.3%, respectively. This method offers a scalable, objective foundation for digital dental diagnostics in addition to automating plaque detection with high fidelity, which may lessen the workload associated with manual evaluation in clinical and research settings.</p>