Image Processing and Temperature Analysis Algorithms for Quality Monitoring of Lettuce Plants

To ensure high-quality growth of lettuce plants, it is essential to acquire the temperature data of the plants and investigate the distribution of canopy temperature. This study focuses on the thermal image processing and temperature analysis algorithms. The SAM (Segment Anything Model) algorithm is used to remove the background in RGB images and segment individual lettuce plants, which are then mapped to CSV files containing temperature data. The temperature of each pixel in the segmented lettuce plant is analyzed, and several indicators, including the mean temperature <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$T_{mean}$</tex>, temperature standard deviation <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$std$</tex>, and coefficient of variation <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$C_{v}$</tex>, are computed. A temperature distribution histogram and kernel density curve are generated to visualize the temperature distribution across the lettuce plant. The temperature standard deviation <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$std$</tex> and coefficient of variation <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$C_{v}$</tex> are utilized to evaluate the quality of the lettuce plants, and K-means clustering is applied to categorize the lettuce plants into three quality levels: high-quality, medium-quality, and low-quality. Experimental results demonstrate that the developed image processing and temperature analysis algorithms efficiently capture the temperature information of lettuce plants, providing valuable technical support for quality assessment and management in precision agriculture.