VarietyQuest: Data-Driven Exploration of Apple Fruit Variety Classification Frameworks

Abstract

The Kashmir valley is known for its breathtaking landscapes and the cultivation of a variety of fruits, with apples being one of the most popular. The Kashmiri apple has a distinct flavor and is widely appreciated for its quality, leading to its export to various destinations around the globe. The sorting and packaging of apples remain a significant challenge due to the lack of skilled labor and the sheer volume of apples produced in the season. As a result, a considerable amount of the harvest is either lost or damaged, leading to significant financial losses. To address this issue, automated classification systems for categorizing fruit varieties using machine learning-based techniques are being developed. Such systems could potentially lead to increased productivity while simultaneously reducing labor costs and errors in sorting and classification. This article provides an overview of the data-driven methodologies for the automated categorization of apple varieties, exploring both conventional and state-of-the-art approaches. The article also provides a concise discussion of the datasets employed in these frameworks. Our study identifies machine learning as a critical foundation for most Apple variety classification frameworks. Further, the absence of datasets for the Kashmiri apple variety is noted during the survey, highlighting the need for further research in this domain. Overall this research explores the automated classification and packaging systems, which can streamline the process and minimize losses while contributing to the growth of the Apple industry across the globe.