Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Recurso digital |
| Language: | English |
| Published: |
Zenodo
2025
|
| Subjects: | |
| Online Access: | https://doi.org/10.5281/zenodo.15430224 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866902078852431872 |
|---|---|
| author | Blessy Harshitha Gandu Sri Harsha Pappu Ruchitha Chebrolu Charan Kumar Marpudi Ashok Kumar Reddy Kaliki |
| author_facet | Blessy Harshitha Gandu Sri Harsha Pappu Ruchitha Chebrolu Charan Kumar Marpudi Ashok Kumar Reddy Kaliki |
| contents | <p><strong>Pharmaceutical Image Recognition Using Deep Learning</strong></p> <h3><strong>Project Description:</strong></h3> <p>This project focuses on the automatic identification of pharmaceutical pills using deep learning and computer vision techniques. It addresses the critical issue of medication errors due to misidentification, particularly when pill labels are damaged or missing. The system is built using <strong>TensorFlow</strong> and <strong>Keras</strong>, employing <strong>Convolutional Neural Networks (CNNs)</strong> to detect and classify pills based on their physical features such as color, shape, and size.</p> <p>A large dataset of pill images—including various types with different angles—is used for training and testing. The project utilizes image pre-processing techniques like <strong>normalization, resizing, and data augmentation</strong>. Advanced object detection algorithms such as <strong>OpenCV’s GrabCut</strong> are applied for precise segmentation.</p> <p>Once a pill image is input via a camera, the model performs real-time identification by matching the image with a pre-trained dataset and retrieves relevant information such as the <strong>pill name, composition, dosage, and use cases</strong>. The system aims to support pharmacists, visually impaired users, and general patients in safe medicine identification.</p> <p>This research also explores the integration of this model into mobile or web applications for broader accessibility and includes performance evaluation with accuracy metrics.</p> <h3><strong>Technologies Used:</strong></h3> <ul> <li> <p><strong>Python</strong></p> </li> <li> <p><strong>TensorFlow</strong></p> </li> <li> <p><strong>Keras</strong></p> </li> <li> <p><strong>OpenCV</strong></p> </li> <li> <p><strong>CNN (Convolutional Neural Network)</strong></p> </li> <li> <p><strong>Image Preprocessing & Augmentation</strong></p> </li> <li> <p><strong>Pillbox / PIRD Dataset</strong></p> </li> </ul> <h3><strong>Key Features:</strong></h3> <ul> <li> <p>Real-time pill recognition via camera input</p> </li> <li> <p>Accurate detection based on shape, color, and size</p> </li> <li> <p>Dataset-driven training with high accuracy (~91%)</p> </li> <li> <p>Dashboard displaying drug details and use cases</p> </li> <li> <p>Potential mobile app integration for broader utility</p> </li> </ul> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_15430224 |
| institution | Zenodo |
| language | eng |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Pharmaceutical Image Recognition Using Deep Learning Blessy Harshitha Gandu Sri Harsha Pappu Ruchitha Chebrolu Charan Kumar Marpudi Ashok Kumar Reddy Kaliki Deep learning Deep Learning Education, Medical Pharmacy <p><strong>Pharmaceutical Image Recognition Using Deep Learning</strong></p> <h3><strong>Project Description:</strong></h3> <p>This project focuses on the automatic identification of pharmaceutical pills using deep learning and computer vision techniques. It addresses the critical issue of medication errors due to misidentification, particularly when pill labels are damaged or missing. The system is built using <strong>TensorFlow</strong> and <strong>Keras</strong>, employing <strong>Convolutional Neural Networks (CNNs)</strong> to detect and classify pills based on their physical features such as color, shape, and size.</p> <p>A large dataset of pill images—including various types with different angles—is used for training and testing. The project utilizes image pre-processing techniques like <strong>normalization, resizing, and data augmentation</strong>. Advanced object detection algorithms such as <strong>OpenCV’s GrabCut</strong> are applied for precise segmentation.</p> <p>Once a pill image is input via a camera, the model performs real-time identification by matching the image with a pre-trained dataset and retrieves relevant information such as the <strong>pill name, composition, dosage, and use cases</strong>. The system aims to support pharmacists, visually impaired users, and general patients in safe medicine identification.</p> <p>This research also explores the integration of this model into mobile or web applications for broader accessibility and includes performance evaluation with accuracy metrics.</p> <h3><strong>Technologies Used:</strong></h3> <ul> <li> <p><strong>Python</strong></p> </li> <li> <p><strong>TensorFlow</strong></p> </li> <li> <p><strong>Keras</strong></p> </li> <li> <p><strong>OpenCV</strong></p> </li> <li> <p><strong>CNN (Convolutional Neural Network)</strong></p> </li> <li> <p><strong>Image Preprocessing & Augmentation</strong></p> </li> <li> <p><strong>Pillbox / PIRD Dataset</strong></p> </li> </ul> <h3><strong>Key Features:</strong></h3> <ul> <li> <p>Real-time pill recognition via camera input</p> </li> <li> <p>Accurate detection based on shape, color, and size</p> </li> <li> <p>Dataset-driven training with high accuracy (~91%)</p> </li> <li> <p>Dashboard displaying drug details and use cases</p> </li> <li> <p>Potential mobile app integration for broader utility</p> </li> </ul> |
| title | Pharmaceutical Image Recognition Using Deep Learning |
| topic | Deep learning Deep Learning Education, Medical Pharmacy |
| url | https://doi.org/10.5281/zenodo.15430224 |