Identification of sunflower seeds with deep convolutional neural networks
- PDF / 1,565,985 Bytes
- 10 Pages / 595.276 x 790.866 pts Page_size
- 67 Downloads / 232 Views
ORIGINAL PAPER
Identification of sunflower seeds with deep convolutional neural networks Ferhat Kurtulmuş1 Received: 17 December 2019 / Accepted: 13 October 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract In the food and agricultural industries, it is crucial to identify and to choose correct sunflower seeds that meet specific requirements. Deep learning and computer vision methods can help identify sunflower seeds. In this study, a computer vision system was proposed, trained, and tested to identify four varieties of sunflower seeds using deep learning methodology and a regular color camera. Image acquisition was carried out under controlled illumination conditions. An image segmentation procedure was employed to reduce the workload in obtaining training images required for training deep convolutional neural network models. Three deep learning architectures, namely AlexNet, GoogleNet, and ResNet, were investigated for identifying sunflower seeds in this study. Different solver types were also evaluated to determine the best deep learning model in terms of both accuracy and training time. About 4800 sunflower seeds were inspected individually for training and testing. The highest classification accuracy (95%) was succeeded with the GoogleNet algorithm. Keywords Sunflower · Seed classification · Deep learning · Neural networks · Computer vision
Introduction Sunflower (Helianthus annulus L.) production in the world is about 47.9 Mt [1]. For some of the countries such as Turkey, sunflower is one of the most important oilseed sources. Sunflower seeds are used for several purposes, such as oil extraction, in confectionery, and as a snack roasted or salted. The seeds have a high nutritional value containing moisture 5.50%, protein 18.72%, crude fat 37.47%, crude fiber 28.30%, ash 3.49% and carbohydrates 6.11% [2]. Sunflower oil is a rich source of unsaturated fatty acids. Many different sunflower varieties have been developed as a result of years of breeding work. It is crucial to choose a correct sunflower seed variety for individual usage purposes and climatic conditions. Food producers and growers want to ensure that the correct variety of sunflowers is processed or cultivated [3–5]. Seed inspection is an important and timeconsuming task in the domains of food and agricultural engineering, especially when production amounts are increased. * Ferhat Kurtulmuş [email protected] 1
Department of Biosystems Engineering, Faculty of Agriculture, Bursa Uludag University, 16059 Bursa, Turkey
For some situations, experts can be consulted to identify raw food materials. However, experts are not always available or affordable when the consultation is needed. Besides, even experts could make mistakes, since humans are prone to errors. Therefore, fast, unbiased, non-destructive, and automated systems for these types of tasks are needed for modern production processes. Computer vision-based methods for identifying food characteristics were reported in various studies. Paliwal et al. [6
Data Loading...
