Overview of image processing approach for nutrient deficiencies detection in Elaeis Guineensis

Abstract

The most common problems occurred in Elaeis Guineensis or widely known as oil palm are plant diseases and pest outbreaks. The diseased oil palm plants normally shows a range of symptoms such as coloured spots or streaks that will occur on the leaves, stems, and seeds of the plant. At present, in the agricultural sectors, diagnosing the type disease of plants are based on human expert, which is alongside with the conventional method applied using test device and performing laboratory test. Therefore, the needs in new approach to classify type of diseases are preferable. Hence, the aim of this paper is to focus on an innovative method based on image processing technique for classifying the lack of nutritional disease occurred in oil palm leaves by analyzing the leave surface only. The result is usable as a guide for fertilization since the trees respond rapidly to the applied fertilizers. The main important concern is to ensure the sufficient amount of fertilizer since excessive intake of fertilizers will cause toxicity to trees and indirectly increase cost of fertilizers. Images of oil palm leaves will be captured using high-end digital imaging device to analyse the leaves surface. Further, feature extraction algorithms also will develop based on shape, texture, and colour of the disease type. The feature vectors will be attained acting as inputs to fuzzy classifier. Overall, the proposed method will benefit the oil palm industries to fulfill the industry demand.