Simon Kasahun Bekele

Simon Kasahun Bekele

Auto-generated profile, based on publicly available information

Simon Kasahun Bekele

Cooperative Bank of Oromia, Addis Ababa, Ethiopia

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
30 June 2025 in Turkish Journal of Engineering
Verified
Journal Article
Open access
PDF file
RJRituraj Jain
Rituraj Jain
SBSimon Bekele
Simon Bekele
DPDamodharan Palaniappan
Damodharan Palaniappan
KPKumar Parmar
Kumar Parmar
PTPremavathi T
Premavathi T
Advances in image processing and techniques in artificial intelligence have made it possible for computers to see and learn. This article introduced a technology that has utilised MobilenetV2 Deep Convolution Neural Network architecture to automatically identify and diagnose plant diseases from images. The identification and classification of plant diseases are now carried out by only human experts-crop extension agents, and farmers, expensive labour that is prone to mistakes. This study relies on dataset gathering as a technique of classifying and identifying plant diseases. It is a multistep process involving pre-process data on the raw set, mask green area of the leaf, remove green section, convert to grayscale and then obtain some characteristics, select, and classify with regard to disease management, etc. Two different types of plants, maize and potato, have been taken in consideration to show effectiveness of the outcome of the proposed model. The confusion matrix and classification performance report were used to evaluate the system. The dataset for potato and maize comprised 6228 and 6878 images, respectively, of leaves. Precise, recall, and F1-scores of 95.15%, 94.76%, and 94.93% were recorded as a cumulative performance across the datasets of potato and maize respectively. This translates to its resistance in picking most diseases for these crops, making it a resource that can be used with confidence in agriculture disease detection. The MobileNetV2 model performs well in both crops, especially for potato early blight and maize common rust. Lower performance in recognizing healthy potato leaves suggests that the feature space of healthy and diseased leaves may overlap. The MobileNetV2 model performed a robust ability in general in the detection of most diseases affecting both potato and maize leaves, but some specific areas need to be targeted for further enhancement.

14 December 2024 in ASEAN Journal of Scientific and Technological Reports
Verified
Journal Article
Open access
PDF file
RJRituraj Jain
Rituraj Jain
SBSimon Kasahun Bekele
Simon Kasahun Bekele
DPDamodharan Palaniappan
Damodharan Palaniappan
KPKumar Parmar
Kumar Parmar
RPRamesh Babu P
Ramesh Babu P
ATAnandkumar M. Trivedi
Anandkumar M. Trivedi
Testing, diagnosis, and treatment must be done quickly in the current scenario, where the COVID-19 pandemic continues to affect the world's health. Pneumonia and Tuberculosis (TB) are always very concerning chest diseases, and COVID-19 has also been added to this list. Chest X-rays (CXR) and CT scans are major sources for diagnosing respiratory disorders. As CT scans are a little bit costlier and take more time, as well as expose patients to mild radiation, the use of CXR is increasing for the diagnosis of chest diseases. This study uses CXR images to accurately identify diseases using the developed MobileNetV2, ResNet50, VGG19, and DenResCov-19 Convolutional Neural Network (CNN). The models were analyzed and tested using the publicly available 13805, 5840, and 4200 chest x-ray images for COVID-19, pneumonia, and TB cases. The in-house mobile app employs a variety of deep-learning models to identify diseases from chest X-rays. With 99.79% accuracy and 96.89% F1-score, ResNet50 is the best model architecture for COVID-19 classification. The ResNet50 pneumonia classification model was 99.17% accurate and had an F1-score of 94.52%. With 99.88% TB detection accuracy and 99.64% F1-score, MobileNetV2 and ResNet50 were equivalent. Comparing model efficacy on the dataset was fascinating.

15 September 2026 in Scientific Reports
Verified
Journal Article
Open access
PDF file
RMRitika Mehra
Ritika Mehra
VTVikas Thapa
Vikas Thapa
RJRituraj Jain
Rituraj Jain
KUKamal Upreti
Kamal Upreti
GPGovind Panwar
Govind Panwar
SBSimon Kasahun Bekele
Simon Kasahun Bekele
Weeds are a major hazard to agricultural production and ecology in the hilly areas of Uttarakhand, India. Traditional weed management methods use blanket spraying of herbicides, which is very expensive, polluting, and unrealistic in steep slope areas where machines cannot reach. This study addresses this issue by recommending a portable robotic platform (robot) with the capability of detecting weeds using a deep learning system to apply precision herbicides to uneven and undulating surfaces. The system utilizes YOLOv11, the most recent version of the YOLO algorithm, which is renowned for its high accuracy and capability to detect objects in real time. It can also distinguish between crops, weeds, and background objects so that they can be sprayed using site-specific spraying. It was compared to Faster R-CNN on a performance basis using conventional detection metrics. On the independent test set, YOLOv11 achieved precision 0.7516, recall 0.8846, F1 0.8127, and mAP@0.5 0.8894, compared with Faster R-CNN at precision 0.3483, recall 0.9538, F1 0.5103, and mAP@0.5 0.8873; five-fold cross-validation confirmed that YOLOv11’s performance was stable across data partitions (mAP@0.5 0.889 ± 0.005, precision 0.750 ± 0.005, recall 0.881 ± 0.006). Although Faster R-CNN had a higher recall rate, YOLOv11 had a higher precision and F1 score, indicating that it is more reliable with fewer false positives. In end-to-end field trials over 0.7 ha of terraced fields, the integrated robot achieved a 92.4% spray hit-rate, a 64.8 ± 3.1% reduction in herbicide volume versus blanket application (mean ± SD; n = 3 independent paired field trials), and 7.1% off-target drift, demonstrating that the title’s claim of autonomous precision spraying for hilly agriculture is operationally validated under the tested field conditions. These results indicate that the proposed system offers a viable pathway toward reduced herbicide consumption, lower operating cost, and reduced environmental impact under hillside field conditions, warranting further multi-season validation.