Fowad Ahmed will present his M.Sc. thesis “Agricultural landscape Configuration and Pattern Analysis with VHR Imagery in Bavaria” on Friday 11th at 11am.
from the abstract: “Hedgerows and woody linear features are an integral part of the landscapes where they exist. These natural and semi-natural physical features deliver several ecosystems services and functions as ecological hotspots to its surrounding flora and fauna. These physical feature mostly found within agriculture landscape are prone to come under increasing pressure, with the human population set to exceed 9 billion by 2050. In this study, a four-band (RGB+Near Infrared) high spatial resolution (1 m) airborne imagery was utilized and following geographic object-based image analysis methodology, the detection of hedgerows on fifteen selected sites across the Bavarian landscape was carried out. The GEOBIA was implemented in ENVI feature extraction module, and SVM algorithm was used for classification purposes. The classification results showed that hedgerows with the combination of VHR imagery and GEOBIA can be identified with good accuracy(Overall accuracy: 0.97%, kappa:0.97). The results highlighted that inconsistency in airborne imagery influences the classification accuracies. A combined land use maps were generated after the inclusion of hedgerows maps within land-use maps from ATKIS DLM and Landklif land use map. Subsequently, the landscape pattern of study sites was analysed through landscape composition and configuration metrics.” supervised by Carina Kuebert-Flock