noImage

작물 상태ㆍ형질도 열영상 카메라로 파악한다
Thermal imaging used to check crop conditions and attributes

Symposium held on plant modeling and AI-based smart farm

Jeju National University’s Subtropical Horticulture Research Institute held an academic symposium with invited specialists on plant modeling and artificial intelligence-based smart farms.

Kim Do-soon, professor at Seoul National University, gave a presentation with the topic of diagnosing stress levels of crops and weeds using the nutrition analysis of plants. Mentioning the use of imaging technologies for medical purposes, Kim discussed using the same non-destructive imaging techniques in agriculture. Data processing takes much shorter time now than ever due to the development of artificial intelligence, he added.

For his study, the researcher conducted an experiment to identify the damage from sea wind caused to canola flowers at four different levels. The damage was difficult to be diagnosed precisely when observed with bare eyes, while the thermal imaging device clearly distinguished the differences.

Kim explained that the flowers under a higher level of sea wind-induced stress showed a rise in its temperature because its stomas became narrower. The experiment proves that using the thermal imaging device helps non-destructively identify the stress imposed on crops, he claimed.

Chung Yong-suk, a Jeju National University professor, said in his presentation that researchers hardly associate plant or crop modeling with smart farms and artificial intelligence. Preceeding studies have been generally led by those with science and technology majors who used the big data obtained from a set greenhouse environment and focused on improving greenhouse facilities, he pointed out. He urged agricultural researchers to take the lead in utilizing botanical imaging technologies to study different plant varieties.

Kim Hak-jin, a biosystem engineering professor at Seoul National University, also suggested in his presentation that thermal imaging helps identify the fresh weight of a crop. He accounted for the process in more detail, saying that a digital camera can measure the area of a lettuce leaf, allowing for the calculation of its fresh weight. Based on the measurement, the amount of vaporization can be predicted to adjust the culture fluid, he added. He also stated that with deep learning and other advanced technologies applied, visual sensing would secure an even greater scalability.