Image repainted method of overlapped leaves for orchid leaf area estimation

Currently, Internet of Things (IoT) has become a hot issue. For agriculture, IoT techniques bring many advantages, such as lowering the time consumed and the manpower required by real time data collection. Taiwan is one of the main orchid exporters in the world, and orchid export accounts for over 70% of the total floral export in Taiwan, and the species of phalaenopsis accounts for 60 % of the floral export. In a previous study, the IoT technique was utilized to automatically measure environmental factors and orchid leaf traits. An automatic image collecting system and an image processing method were employed to calculate the leaf area and the growth rate of the leaf area which was positively related to the blooming quality. However, when calculating the leaf area, the overlapping leaves caused inaccurate results. To solve this problem, this paper proposes a method to repaint the overlapped area. The leaves of orchids are symmetrical, and this feature is utilized to obtain the overlapped area and repaint the area. Finally, two data sets of processed images are collected to verify the proposed method. This paper provides an effective method of repainting overlapped leaf areas. Using the proposed method and the leaf area estimation method can reduce the error caused by leaf overlapping and increase the accuracy of leaf area estimation and growth rate calculation.