Towards an automated plant height measurement and tiller segmentation of rice crops using image processing

• Main Authors: Constantino, Karol Paulette, Gonzales, Elisha Jeremy, Lazaro, Lordd Michael N., Serrano, Ellen Chelsea, Samson, Briane Paul V.
• Format: text
• Published: Animo Repository 2015
• Subjects: Image processing; Rice; Phenotype; Software Engineering
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/3432
• Institution: De La Salle University

Plant phenotyping is the process of completely assessing the basic and complex characteristics of the plant, which includes height and tiller count. The International Rice Research Institute (IRRI) researchers does plant phenotyping to observe changes in the physical characteristics of the C4 rice crops after modifying its genetic makeup to increase yields without using too much water, land and fertilizer resources. As this advances, the traditional way of observing phenotypic data is still trailing behind. Automated plant phenotyping offers an effective substitute because it allows a regulated image analysis that can be reproduced due to the automation. This is to address the lack in accuracy, reproducibility and traceability in manual phenotyping. With this, an image processing system that automates the measuring of height and the counting of tillers of a rice crop, specifically the C4 rice, was developed. The system applies HSV and Thresholding for preprocessing, Canny Edge Detection (tiller) and Zhang-Suen Thinning Algorithm (height) for the plant structure and tracing and conversion for measuring the height. Tiller counting is done by counting the cluster of pixels in a given region of interest. Four experiments were conducted using different setups and different combinations of algorithms. The fourth experiment was able to get an average percentage error of 76.14% for the tiller count and 238.11% for the height measurement. Presence of shadows and hanging leaves heavily affected the results of this experiment. © 2015, Mechatronics and Machine Vision in Practice. All rights reserved.