Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Chine → 2009 → The Research on Remote Sensing Image Segmentation Technology of Ulan Buh Desert in Inner Mongolia

Northeast Forestry University (2009)

The Research on Remote Sensing Image Segmentation Technology of Ulan Buh Desert in Inner Mongolia

王鹤智; Wang He Zhi

Titre : The Research on Remote Sensing Image Segmentation Technology of Ulan Buh Desert in Inner Mongolia

Auteur : 王鹤智; Wang He Zhi

Grade : Master’s Theses 2009

Université : Northeast Forestry University

Résumé
With the rapid development of remote sensing technology, gaining information through the remote sensing images, such as Ulan Buh Desert of high input, low efficiency, difficult operation and bad environment ,it is important research on the investigation. Due to the complexity of characteristic information of remote sensing image, there is no universal relevance method for image segmentation.In this paper, Ulan Buh Desert color aerial remote sensing image as the main study. By the method of image segmentation, mathematical morphology and choice of threshold segmentation methods, using Matlab 7.0 software makes segmentation and edge detection. In the segmentation process, a typical representative of the desert region and rare in nature of cluster-like sand dunes Nitraria as main study. Using mathematical morphology theory, it could air from the color image segmentation by the photos, and also use this method of vegetation on the ground to image segmentation. Through the shape comparison between thorns colors segmentation map and field survey results, its segmentation effect is better, and also through the ultimate way to verify the measurement accuracy with high precision split gill dunes. In the edge detection process, the use of threshold segmentation and mathematical morphology operation of remote sensing images for edge detection features. By contrasting the edge with the original image, we can see the extent of their match, and also take the contrast gray-scale extraction methods to verify its effect is good. Though the research on Ulan Buh Desert remote sensing image segmentation techniques, it aims to provide technique reference to the vegetation survey of desert-like region by using remote sensing image segmentation

Mots clés : Mathematical morphology image segmentation remote sensing image threshold segmentation edge detection

Présentation (CNKI)

Page publiée le 22 septembre 2014, mise à jour le 19 février 2018