<![CDATA[Chili is one of the main agricultural commodities in Indonesia with significant economic value. Various types of chili, such as large chili, bird's eye chili, and green chili, are often difficult to distinguish manually due to their physical similarities. To support advancements in the agricultural sector, this study utilizes artificial intelligence technology, specifically the Convolutional Neural Network (CNN) with VGG-16 architecture, to automatically identify chili types through image analysis based on color and shape. This study aims to measure the accuracy, sensitivity, and specificity of the model in classifying chili types. The results show that the VGG-16 architecture achieved 100% accuracy in training data testing, indicating the model’s ability to detect and classify chili types optimally. In the model evaluation (fold 5), the accuracy was 91.8%, sensitivity was 88%, and specificity was 93.8%. This study confirms that CNN with VGG-16 is effective for image classification, especially when test data shares similar characteristics with training data. This system offers significant potential for application in the agricultural sector, particularly in improving the efficiency and accuracy of identifying other agricultural commodities.]]>
Copyrights © 2024
