Comparison of object classification methods in seed stream separation

Comparison of object classification methods in seed stream separation

Λεπτομέρειες βιβλιογραφικής εγγραφής
Parent link:Advances in Computer Science Research
Vol. 72 : Information technologies in Science, Management, Social sphere and Medicine (ITSMSSM 2017).— 2017.— [P. 179-181]
Κύριος συγγραφέας: Vlasov A. V. Andrey Vladimirovich
Συγγραφή απο Οργανισμό/Αρχή: Национальный исследовательский Томский политехнический университет Инженерная школа информационных технологий и робототехники Отделение информационных технологий
Άλλοι συγγραφείς: Fadeev A. S. Aleksandr Sergeevich
Περίληψη:Title screen
The paper presents a study of machine learning approaches to detect and classify seeds of a grain crop in order to enhance agricultural seed purification line. The main features of seeds that are hard to recognize during a separation with mechanical methods are resolved with the help of machine learning approach. The main machine learning methods used in research was traditional machine learning and deep learning based on neural networks. A special training image database was retrieved in order to check if the stated approaches are reasonable to use and develop. A set of tests is provided to show the effectiveness of the machine learning applied to solve the stated problem.
Έκδοση: 2017
Θέματα:
электронный ресурс
труды учёных ТПУ
image processing
seeds sorting
classification
feature extraction
convolutional neural network
automatic detection
grains
agriculture
обработка изображений
классификация
нейронные сети
автоматическое обнаружение
зерна
сельское хозяйство
машинное обучение
зерновые культуры
Διαθέσιμο Online:http://dx.doi.org/10.2991/itsmssm-17.2017.38
Μορφή: Ηλεκτρονική πηγή Κεφάλαιο βιβλίου
KOHA link:https://koha.lib.tpu.ru/cgi-bin/koha/opac-detail.pl?biblionumber=657526
Περιγραφή
Περίληψη:Title screen
The paper presents a study of machine learning approaches to detect and classify seeds of a grain crop in order to enhance agricultural seed purification line. The main features of seeds that are hard to recognize during a separation with mechanical methods are resolved with the help of machine learning approach. The main machine learning methods used in research was traditional machine learning and deep learning based on neural networks. A special training image database was retrieved in order to check if the stated approaches are reasonable to use and develop. A set of tests is provided to show the effectiveness of the machine learning applied to solve the stated problem.
DOI:10.2991/itsmssm-17.2017.38

Παρόμοια τεκμήρια