Evaluation of hemp seedling size using ImageJ software

In vitro hemp seed germination is crucial for investigating factors impacting production conditions. It is laborious and time-consuming to measure and evaluate the morphological traits of seedlings grown in vitro. One of the most well-known machine vision techniques is image processing, which offers detailed information and more reliability and accuracy than traditional counting and visual determination of seedlings. The study aimed to determine if hypocotyl images obtained with semi-automatic processing through SmartRoot could be replaced with faster, automatic processing. Seedlings were scanned in RGB color profile in 300 dpi, using an open lid flatbed scanner HP Scanjet G4050. The images were processed in ImageJ software. The average cross-correlation coefficient between hypocotyl images was 0.86. The average diameter and length with the automatic process were 0.111 and 0.865, respectively, and with SmartRoot 0.107 and 0.877, respectively. Using a t-test it was concluded that the average length and diameter obtained with the two methods didn’t show a statistically significant difference. Comparing length and diameter acquired from the two methods showed a correlation of 0.989 and 0.973, respectively. Results showed that hypocotyl images obtained with the automatic process can replace the ones with the SmartRoot process. Macro in ImageJ could be recorded, and images could be batch processed, reducing time for analysis.