Redefining Raisin Quality Control through Advanced Sorting and Visualization with Deep Learning Models

This study introduces a new approach for classifying and displaying raisins by utilising a Deep Learning-Based Support Vector Machine (SVM) model. The processing of raisins, which is a crucial aspect of the agricultural industry, has traditionally depended on labor-intensive and error-prone manual sorting methods. The proposed methodology utilises deep learning techniques to enhance the accuracy and efficiency of this procedure. The Raisins dataset consists of two classes, Besni and Kecimen, which are retrieved from the Kaggle open source platform. The dataset is pre-processed using the SVM model, resulting in a model that can predict with an accuracy of $\mathbf{9 2 \%}$. Raisins have a long and significant cultural heritage that dates back to ancient times. They are employed in several culinary contexts, including savoury dishes and traditional sweets. Raisins are rich in vital nutrients such as vitamins, minerals, and dietary fibre. Due to their inherent sweetness, they can serve as a more health-conscious substitute for refined sugars in a variety of culinary preparations. This study explores the application of a deep learning-based Support Vector Machine (SVM) model in the field of food processing to classify and arrange raisins. The production of raisins involves intricate quality control processes that formerly depended on labor-intensive methods. This work aims to enhance the accuracy and efficiency of sorting raisins by automating their categorization based on quality characteristics. This is achieved by employing Support Vector Machines (SVMs). The SVM model, renowned for its robustness in classification applications, is employed to detect complex patterns in sultana samples, hence improving sorting precision. Moreover, the research explores the visual aspect, shedding light on the structural characteristics of raisins and providing useful insights on quality standards.