Home   /   Contact us

فارسی   /   Mobile

Saturday 15 November 2025




Pejman Tahmasebi

Associate Professor

Natural Engineering

Faculty of Natural Resources and Geosciences

Tel: 038332324401-4

pejman.tahmasebi@sku.ac.ir

Click to view the complete profile


Research Interest
   1- "AI and Machine Learning for Integrated Natural Ecosystem Health Monitoring": This research develops AI-driven solutions to assess ecosystem health through integrated analysis of multi-source data (mobile, drone, and satellite imagery). Using computer vision (OpenCV, PyTorch) and machine learning (CNNs, ViTs, scikit-learn), we extract key ecological indicators like vegetation cover, biomass, and soil health. Our pipeline combines high-resolution drone data for local precision with satellite imagery for broad-scale trends, enabling scalable habitat monitoring. Deep learning models automate species identification and biomass estimation, while fusion techniques correlate ground-truth data with remote sensing. The system delivers cost-effective, real-time insights through deployable dashboards, supporting evidence-based conservation. By bridging field ecology with AI, we create adaptive tools for biodiversity preservation and climate resilience. This interdisciplinary approach advances precision ecology through innovations in data fusion, anomaly detection, and predictive modeling for sustainable ecosystem management.
   2- "Automated Plant Species Identification Using Deep Learning and Image Processing": This research focuses on developing an automated plant species identification system using advanced deep learning and image processing techniques. The study employs computer vision algorithms and deep learning models including Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) for accurate classification of plant images. A comprehensive database of field images of various plant species will be collected and processed for model training and evaluation. The designed system can automatically identify species with high accuracy and speed, and can be utilized in biodiversity studies, environmental monitoring, and ecological research. This technology effectively addresses the challenges of manual plant identification such as human error and time-consuming processes. The system will serve as an efficient tool for botanists and environmental researchers. The development of this technology represents a significant step toward smartening floristic studies and biodiversity conservation efforts.

Academic Resume

Publications



Books



Honors









All rights belong to Shahrekord University