Most of the people who live in urban areas have an increasing need for food. One of the activities that can utilize land and resources in urban areas is urban farming. One example is the Sungkai Green Park Ecotourism in Nagari Lambung Bukit, Pauh Padang District, which has various types of plants that are very useful. However, due to the many and varied plants, farmers experience difficulties in terms of plant maintenance. Therefore, this journal aims to classify weeds and plants, in order to facilitate farmers in plant care. The method that will be applied uses machine learning with image processing to separate weeds and plants which can reduce manual work visually. Image analysis accurately detects areas of identified weeds. Each image has a different pattern and spatial distribution and can be detected using the GLCM technique. This technique represents the relationship between two conflicting pixels in the image. Overall, the image is taken using a webcam which is positioned vertically against the sample plant at low brightness conditions so that it can be processed accurately in machine learning. This GLCM feature is able to extract images to separate weeds and plants by using a texture matrix from the image. The output of the matrix in the form of parameters mean, skewness, kurtosis, entropy, contrast, and energy. These parameters are used to obtain a value for plants so that weeds can be distinguished from plants. The results of this journal can provide early information to farmers to immediately carry out plant maintenance.
Copyrights © 2022