Algoritma pengolahan citra untuk deteksi jeruk lemon (Citrus medica) menggunakan kamera online

Abstract

This research aimed to supporf the application of automatic lemon harvesting using camera. Detection of lemon existence by means of image pmssing started on the basis of color signal analysis of the surface of an object and its backgmund. To realize this idea, some images of Iemon fruits aged f 20 day after flowering were mcorded. The image data were pwssed to get the RGB information. Furthemom, the data were derived to get RGB index and HSI model for deep analysis required. The analysis was done in order to get some parameters that could be used to separate fmB from their backgtvund. Based on the experiment, separation of lemon fruits from their background including leaves, stems, flows, and others could be done by the use of R > 200, G > f 80, and R=G=B < 248 as main farmula and 2r- 0%-b 2 0. f 5 and 2r-0.59-b S 0.55 as an additional formula at /eve! of illumination of 40 to 60 lux. Result of algorithm tested using the parameters showed that crucial factor implying to the unstable result was illuminance.