Segmentasi Citra Hama Daun Tumbuhan Kelapa Sawit Untuk Menentukan Populasi Telur Ulat Menggunakan Algoritma Watershed

Pertiwi, Hanafia (2022) Segmentasi Citra Hama Daun Tumbuhan Kelapa Sawit Untuk Menentukan Populasi Telur Ulat Menggunakan Algoritma Watershed. Other thesis, Universitas Islam Riau.

Text 163510350.pdf - Submitted Version Download (2MB)

Abstract

The pests that most damage oil palm plantations are caterpillars that eat oil palm leaves, namely fire caterpillars and bagworms. damage caused by caterpillars resulting in production reaching up to 35%. Meanwhile, the damage caused by the pouches resulted in up to 40% productivity. Prevention of breeding caterpillar pests is carried out by the HPT (Pests and Plants Diseases) team manually. The HPT team must take samples first for pest sampling must perform the stages of preparing the schedule of early observation implementation, establishing sample points and sample lines, and determining the sample principal. It will take a long time to get results. To speed up the detection time of the population of pest caterpillar eggs, segmentation is carried out using digital image processing with steps consisting of preprocessing and segmentation. The prepocessing process consists of image acquisition, cropping, HSV (Hue, Value, Saturation) and binary color image, followed by distance transform segmentation process, watershed algorithm and watershed marker to separate stacked pest caterpillar eggs. The result is that by testing the credibility of the system, the accuracy value obtained is 89.39% and the average percentage is obtained by 94%. This shows that the system has sufficient ability to know the egg population of pest caterpillars.

Actions (login required)

View Item