Klasifikasi Penyakit Pada Citra Daun Sayuran Kangkung Berdasarkan Fitur Warna Menggunakan Algoritma KNN

Hasdimeyra, Edrian (2023) Klasifikasi Penyakit Pada Citra Daun Sayuran Kangkung Berdasarkan Fitur Warna Menggunakan Algoritma KNN. Other thesis, Universitas Islam Riau.

Text 193510033.pdf - Submitted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

In Indonesia, kale is the vegetable most often consumed by the public, based on survey data from the Central Statistics Agency for 2019 to 2022. Kale is the most popularly consumed compared to other vegetables, and even based on observations of farmers, kale is a vegetable that is easy to grow and maintain until it is sold to the market. farmers experience problems because kale leaves are easily affected by disease and the disease easily spreads from one leaf to another, therefore it is necessary to find out the disease earlier so that it can prevent the disease from spreading more widely, one of which is with the changes that occur in kale leaves. This study tried to detect leaf diseases of kale, namely leaf spots, rotten leaves, and yellowing leaves. The image classification process is carried out through image acquisition, preprocessing and then image segmentation using grabcut to separate the background and images of kale leaves, then the hsv and rgb image values are taken for storage and compared with the hsv and rgb image values in the database using the K - Nearest Neighbor algorithm ( KNN) with the shortest distance. Of the 120 training image data and 20 test image data using the RGB-HSV-KNN algorithm, the best accuracy results are 95% using a k value of 5.

Actions (login required)

View Item