دانشگاه تبریز نشریه مکانیزاسیون کشاورزی 2383-126X 7 1 2022 03 21 Assessment of uniformity of corn seedling planting in transplanted using image processing ارزیابی یکنواختی کاشت نشاء ذرت توسط دستگاه نشاءکار با روش پردازش تصویر 23 33 14947 10.22034/jam.2022.14947 FA سامان مهدی زاده دانشکده مهندسی زراعی و عمران روستایی، گروه مکانیک بیوسیستم، دانشگاه علوم کشاورزی و منابع طبیعی خوزستان هادی اورک دانشکده مهندسی زراعی و عمران روستایی، گروه مکانیک بیوسیستم، دانشگاه علوم کشاورزی و منابع طبیعی خوزستان فاطمه کاظمی کرجی دانشکده مهندسی‌ زراعی و عمران روستایی، گروه مکانیک بیوسیستم، دانشگاه علوم کشاورزی و منابع طبیعی خوزستان Journal Article 2022 07 16 Abstract Cultivation of corn as a seedling could be considered as a way to produce an early crop, reduce production time and also save water and agricultural inputs. Therefore, the aim of the present study is developing an algorithm for evaluation of the planting uniformity in transplanting using image processing. In order to determine the number of seedlings, different transformation methods (Excessive green, green and red normalized difference index and green and blue normalized difference index) were evaluated. Then, the corn seedlings information was taken out from the circumstantial information using Otsu’s method. Finally, by applying Freeman labeling and Harris method and analyzing the characteristic parameters of the overlapping regions of the fields, an automatic counting of corn seedlings in the overlapping regions was established. The developed algorithm was evaluated online at speeds 5, 8, 10 and 12 km / h. According to the obtained results, the best uniformity and single transplanting were achieved at a speed of 5 km / h with an accuracy of 97%. Furthermore, the highest percentage of incorrect planting was obtained 21% at a speed of 12 km / h. Therefore, the study demonstrated that the newly developed algorithm is reliable not only for automatic corn seedlings counting in the corn field but also for the uniformity evaluation of the transplanter. Abstract Cultivation of corn in the form of seedlings can be considered as a solution for producing early crops, reducing production time, and saving water and various agricultural inputs. Therefore, the aim of the current research was to develop an algorithm to evaluate the uniformity of planting in the nursery with the help of image processing. In order to determine the number of seedlings, various transformations (excessive greenness, normalized intensity difference index of red and green values and normalized intensity difference index of blue and green values) were evaluated. Then, by applying thresholding by Etsu method, the information of corn seedlings was separated from the background. Finally, by implementing the Freeman labeling method and the Harris method and analyzing the features of overlapping areas, an automatic algorithm was developed to count corn seedlings in these areas. The developed algorithm was evaluated online at speeds of 5, 8, 10 and 12 km/h. According to the obtained results, the best uniformity and monoculture was obtained by the transplanter at a speed of 5 km/h with a percentage of correct cultivation of 97% and the highest percentage of incorrect cultivation at a speed of 12 km/h and equal to 21%. Therefore, based on the present study, according to the obtained accuracy, the developed algorithm can be used not only to count the number of corn seedlings in the field, but also to evaluate the uniformity of planting of seedlings.

https://jam.tabrizu.ac.ir/article_14947_268da9433d21a97594a0a50dc42ead9a.pdf