Quality Assessment of Turfgrasses Using NTEP Method Compared to an Image-Based Scoring System

The current methods of turfgrass evaluations are often based on human-based assessment methods. However, eliminating subjective errors from such evaluations is often impossible. This research compared the accuracy of human-based and digital image processing-based methods for quality assessment of turfgrasses. Four turfgrass plots were evaluated using the two mentioned methods. In the human-based method, 20 evaluators (10 women and 10 men) and in the image-based method, a digital camera with an artificial and controlled light source were used. This experiment for the first time evaluated the two qualitative characteristics of turfgrass texture and weed growth tolerance using a specific image processing-based technique and the common human-based evaluation method. Further, total coverage, color, and living coverage of the turfgrasses were compared with the two methods. The results of the human-based assessment method showed a wider range and higher standard deviations than that in the image processing method, which seems to be due to the differences between the human's evaluators and errors caused by the human mind. The results also emphasized the accuracy and ease of application of the image-processing-based method. This outcome can have applications for developing a mechanized system for turfgrass quality evaluation across the world.