Chili Disease Detection Using HOG with Euclidean Distance

Abstract

In order to detect plant diseases in the leaves of chili plants, automatic learning is used in this study. Farmers are planting chilies with the intention of exporting them worldwide. Chili is a need for regular meals. There aren't many illnesses that need to be found in the leaves of chili plants. There are three types of chili plants: weak, diseased, and healthy. Weak and sick chili plants can be affected by diseases such as a harsh leaf, spot leaf, whitefly, yellowish, etc. It has been reported that research is underway to determine whether chile plants are safe to grow or polluted. But when it comes to agriculture, it's critical to recognize the damaged plant by its unique type. Various category diseases are studied using the HOG (Histogram of Oriented Gradients) of the leaf of the chili plant. The representative feature vectors in the feature vector are created using the mean value of every feature point. A typical feature vector and the Euclidean distance are used to calculate the outliers. For the Euclidean distance larger than 0.0025, 0.0016, and 0.00125, the average accuracy rate was 61.6%, 73.2%, and 81.00%, respectively, with the modified border point in the feature vector being 0.0016, 0.00125, and 0.0009. The results presented above suggest that machine-learning techniques for image processing can be used to determine the type of plant disease.