Maize seed variety identification model using image processing and deep learning
- Authors: Gebeyehu, Seffi , Shibeshi, Zelalem S
- Date: 2024
- Language: English
- Type: text , article
- Identifier: http://hdl.handle.net/10962/429191 , vital:72566 , DOI: 10.11591/ijeecs.v33.i2.pp990-9985
- Description: Maize is Ethiopia’s dominant cereal crop regarding area coverage and production level. There are different varieties of maize in Ethiopia. Maize varieties are classified based on morphological features such as shape and size. Due to the nature of maize seed and its rotation variant, studies are still needed to identify Ethiopian maize seed varieties. With expert eyes, identification of maize seed varieties is difficult due to their similar morphological features and visual similarities. We proposed a hybrid feature-based maize variety identification model to solve this problem. For training and testing the model, images of each maize variety were collected from the adet agriculture and research center (AARC), Ethiopia. A multi-class support vector machine (MCSVM) classifier was employed on a hybrid of handcrafted (ie, gabor and histogram of oriented gradients) and convolutional neural network (CNN)-based feature selection techniques and achieved an overall classification accuracy of 99%.
- Full Text:
- Date Issued: 2024
Information extraction model from Ge’ez texts
- Authors: Gebeyehu, Seffi , Wolde, Worke , Shibeshi, Zelalem S
- Date: 2023
- Language: English
- Type: text , article
- Identifier: http://hdl.handle.net/10962/429176 , vital:72565 , 10.11591/ijeecs.v30.i2.pp787-795
- Description: Nowadays, voluminous and unstructured textual data is found on the Internet that could provide varied valuable information for different institutions such as health care, business-related, training, religion, culture, and history, among others. A such alarming growth of unstructured data fosters the need for various methods and techniques to extract valuable information from unstructured data. However, exploring helpful information to satisfy the needs of the stakeholders becomes a problem due to information overload via the internet. This paper, therefore, presents an effective model for extracting named entities from Ge'ez text using deep learning algorithms. A data set with a total of 5,270 sentences were used for training and testing purposes. Two experimental setups, i.e., long short-term memory (LSTM) and bidirectional long short-term memory (Bi-LSTM) were used to make an empirical evaluation with training and a testing split ratio of 80% to 20%, respectively. Experimental results showed that the proposed model could be a practical solution for building information extraction (IE) systems using Bi-LSTM, reaching a training, validation, and testing accuracy as high as 98.59%, 97.96%, and 96.21%, respectively. The performance evaluation results reflect a promising performance of the model compared with resource-rich languages such as English.
- Full Text:
- Date Issued: 2023