DefoM
| #Web Application #Computer Vision #Satellite Image Processing #MLOps
Description
Our project, DefoM (Deforestation Tracking and Monitoring), is a comprehensive application that amalgamates cutting-edge AI technology with a mobile and web interface. My primary role in this endeavor involved spearheading the development of the AI model and API. The AI component of our system centers on image segmentation, with a focus on tracking changes in forested areas by analyzing satellite imagery. Additionally, we created a classification model to identify specific forest vegetation. The project also involved the orchestration of a complete pipeline for daily satellite image extraction and annotation using deep learning models, with data storage in MongoDB. DefoM represents a robust tool for deforestation monitoring.
Features
- AI Model Integration: DefoM incorporates AI models for image segmentation and classification, enabling it to effectively track and monitor changes in forested areas.
- Satellite Image Analysis: Our project performs in-depth analysis of satellite imagery, allowing it to detect deforestation and changes in forest vegetation.
- Data Pipeline: We implemented a full data pipeline to handle the extraction of daily satellite images and annotation using deep learning models.
Technologies
Methodologies
Image Segmentation Classification Model Data Extraction Data Lakes
Lessons Learned
- Complexity of Remote Sensing Data: Dealing with satellite imagery and remote sensing data comes with its unique set of complexities, particularly in terms of preprocessing and model training.
- Data Management: The project underscored the importance of effective data management, especially when dealing with extensive satellite datasets.
- Cross-Platform Integration: Integrating AI models with both mobile and web applications required a well-thought-out strategy and coordination.