Performance Analyses of Various Kernel Function Ml Techniques in Groundnut Seed Classification

Groundnut oil is a commodity widely consumed throughout the world, with its quality directly influenced by the nature of the seeds. Traditional manual inspection techniques are laborious and prone to human error, creating a need for automated classification methods. This work focuses on using Support Vector Machines (SVMs) with polynomial, sigmoid, and Laplacian kernels in classifying groundnut seeds. A comprehensive dataset of groundnut seed images was preprocessed, and key features such as texture, shape, and color were extracted using advanced image processing techniques. The Laplacian kernel outperformed the others, achieving the highest accuracy of 92% and the shortest computation time, demonstrating its suitability for real-time applications. Selection of kernels in SVMs for agricultural application: This paper draws attention to the importance of kernel selection in SVMs towards improving the efficiency of seed classification systems.