Image Processing System for Early Detection of Cocoa Fruit Pest Attack

Abstract

Cocoa (Theobroma cacao L.) is one of the leading commodities that developed in quality and quantity. Efforts through socialization and counseling in the field of cocoa cultivation for the farmers are always reminding know the importance of identifying symptoms from pest attacks could earlier makeup so that preventive actions that do not damage the environment or agricultural land. The study aims to create an early detection system based on image processing on symptoms of pest attacks on cocoa fruits. The early detection system involves training the data and test the data as a result of capture pictures of ordinary cocoa fruits and reviews those attacked by pests in real time at the location of cocoa plantations. Image processing techniques are integrated into the application software to be Able to identify the pixel characteristics of capture images of cocoa fruits. The results of the study showed that the ability of the system to detect the symptoms of pests in training the data was 100% and the test of data was 70%. This result showed that the applications could be recommended to be developed on a larger scale so that it will be helpful for cocoa farmers.