Publication

Abstract : In India, there have been many technological breakthroughs in every area, yet there is still a significant lack of technological advancements in agriculture. A computer vision-based model will be suggested in this study to measure and estimate the diameter of a coconut tree trunk because the traditional technique is manual and requires each tree to be measured individually, making the process laborious and time-consuming. Furthermore, due to the difficulty of the experiment, manual measurements cannot be performed regularly. Obtaining the diameter metric of a tree using computer vision will have numerous uses in the sectors of commercial forestry, ecology, horticulture, and it may be infused into current autonomous coconut robots. It is critical to understand the precise measures of a coconut tree trunk since this parameter will aid in the effective management of coconut trees as well as the identification of their health. This study evaluates the three existing computer vision-based object size measurement algorithms on coconut tree trunks. The algorithms that are implemented in this paper are using simple open cv methods including canny edge, erosion, dilution, Gaussian filter, and contours. We will provide results by comparing and evaluating the efficacy and accuracies of existing object size measuring algorithms on coconut tree trunks on a range of trunk images acquired directly from the real world, by quantifying the performance using ground truth on a subset of those images.