Conceptual framework of smart fertilization management for oil palm tree based on IOT and deep learning

Abstract

Oil palm trees contribute economic income to the national and community by generating various types of productions. This will cause an expansion of the area for the plantation of oil palm seeds, then contributes to the stability in distributing good quality oil to accommodate the growing population. Furthermore, degradation occurs when the planting of oil palm trees increases rapidly, especially the occurrence of uncontrolled oil palm cultivation. The degradation can cause loss of soil nutrients due to soil erosion. The lack of macronutrients, Nitrogen (N), Phosphorus (P), Potassium (K) and Magnesium (Mg) on oil palm tree may impact on its growth which includes the quality of crops. Traditional approach to detect macronutrients, can also lead to some improper control in turn leads results in reduction in yield. The existing system has given limited information of dataset and slower classification performance due to limited functions. With the adaptability of Internet of Things (IoT) technologies, oil palm tree growth data and fertilization management can be utilizing effortlessly and effectively. The context of conceptual framework comprises the IoT technologies, image processing, machine learning and deep learning which focuses on environmental factors that affecting the young oil palm tree growth that involve temperature, humidity, soil moisture content, light and nutrient. Thus, a study of IoT, machine learning and deep learning for smart fertilization management of oil palm trees is suggested helping and raise the efficiency of oil palm trees management in Malaysia.