A deep learning-based seed classification with mobile and web application

Seed quality is an essential factor in agricultural production. In the recent years, advanced technologies are needed to increase productivity in the agricultural sector. The contribution of artificial intelligence studies in the agricultural sector has emerged because of this need. Some seeds are small in nature and it is difficult to identify and classify differences between species. In the traditional method, it is decided by experts to define and classify these differences, considering the morphological structure, shape, texture and color. This method involves a classification process that is costly, subjective and time confusing, what makes it necessary to develop a process that can automatically detect the type of seeds. In this study, a mobile application and web application has been developed that quickly detects and classifies seed images with high accuracy using CNN, one of the deep learning techniques. The training process was carried out by creating a dataset consisting of 15 widely known seed images. InceptionV3, Xception and InceptionResNetV2 models achieved 99% accuracy, and ResNet50 model 98% accuracy. The model, which works with high accuracy has been enabled to work in mobile and web environments. In addition, thanks to the mobile application, detailed information about the morphological characteristics and use of the seed identified was provided to the user.