TY - JOUR AU - taghimollaei, yousef AU - Karamshahi, Abdolali PY - 2019/01/30 Y2 - 2024/03/29 TI - Vegetation mapping through using satellite images of ‎WorldView 2- A case study of Haft Barm, Shiraz ‎ JF - Journal of Wildlife and Biodiversity JA - JWB VL - 3 IS - 1 SE - Original Article DO - 10.22120/jwb.2018.88578.1027 UR - https://wildlife-biodiversity.com/index.php/jwb/article/view/107 SP - 9-21 AB - <p class="justify_txt">Land cover maps are regarded one of the main inputs for land use planning and environmental modeling. One of the main reasons of unsuitable spatial array of the urban areas can be related to the rural communities' emigration, which in turn cause complete degradation of the farmlands and rural structures. Such event can be regards as a factor which is responsible for demolishing of the Haft Barm area an important recreational and touristic areas in&nbsp; the vicinity of Shiraz metropolis. Therefore, recognizing the natural conditions of the area, preparing resource maps like land use and land cover, and monitoring their changes during the time is critical issue in the&nbsp; environmental planning and management. To this aim, WorldView 2 images with eight bands were used and&nbsp; mentioned maps were produced. The mapping analysis way was relied on an object-based classification methodology and using a decision tree which was applied in the WorldView 2 images categorization. The process shall be as the following: a) segmentation, b) terrain selecting, c) creating a decision tree for images' classification, and d) ultimate classification and evaluation of the accuracy. The area was divided into 10 user classes. The results indicated successful classes categorization with overall accuracy of 87.45%. The highest accuracy of classification was obtained for water, forest, product, building classes respectively. Planted forests patches as well as natural forests were identified and classified using OBC approach (object-based method) while additional coastal bands were used to distinguishing among barren and covered lands. Distance to tree and shadow play an important role in identifying buildings.</p> ER -