利用光譜訊息來改進多層次型態學動態輪廓演算法於樹木偵測與樹冠描繪

Improving Multi-level Morphological Active Contour for Tree Crowns Detection and Delineation using spectral information

陳怡玲(Yi-Ling Chen)

指導教授 ： 吳昭正

國立臺北科技大學/電資學院/電機工程系研究所/碩士(2013年)

https://doi.org/10.6841/NTUT.2013.00348

摘要

台灣的森林由於氣候與海拔的關係，從低海拔的闊葉林到高海拔的針葉林，分佈相當廣泛。台灣的高山地形使得人工管理近乎不可能，因此遙測影像被廣泛的應用來解決此一問題，然而人工影像判讀除了需要大量的人力以外，並嚴重依賴判讀者的經驗。近年來多層次型態學動態輪廓演算法(MMAC)被提出來用以自動偵測樹木與輪廓描繪。
多層次型態學動態輪廓演算法(MMAC)雖然可以有效的解決高山地區的樹木輪廓辨識率，然而此一演算法的原先設計，是針對光達影像所提供的高度資訊來偵測樹木與樹冠描繪。若影像資料為傳統的RBG影像與多光譜影像的型態，因為難以直接對應到光達影像所呈現的高度時，此一演算法的偵測效能就會大幅下降。
本論文將光譜資訊加入多層次型態學動態輪廓演算法，用以改良其原本只能使用於光達影像上的缺點，藉由加入頻譜資訊的概念，利用頻譜分析與分類器，進一步提升偵測率並拓展此一演算法到多光譜影像。本論文以光譜分析的角度改進原演算法的限制性，並進一步提升樹木偵測與樹冠描繪的效能，並期望未來能夠應用於大面積的遙測影像上。

關鍵字

樹木偵測 ； 樹冠描繪 ； 多層次型態學動態輪廓演算法 ； 光譜分析 ； 分類器

並列摘要

Forests in Taiwan distribute vertically along the central region and can be categorized into broadleaved, mixed, and conifer forests. Terrain features make manual inspection of forests nearly impossible. By utilizing remote sensing data, the amount of field sampling could be significantly reduced. However, the visual interpretation is labor-intensive and heavily dependent on the interpreter’s experience. A new algorithm, called multi-level morphological active contour algorithm (MMAC), has been proposed by Prof. Lin, to address these issues in 2011.
The MMAC could effectively increase recognition rate of individual tree in mountainous areas, which is the common case in Taiwan. However, the design of algorithm could only cope with Lidar images, which contains altitude information of ground objects. If the RGB or multispectral images were applied, the performance of MMAC would drop significantly.
This thesis exploits spectral information to eliminate restriction of MMAC which only runs on lidar images. Multispectral analysis and classifiers would be exploited to provide spectral features and classification of tree tops and their crowns. The contribution of this thesis would extend the applications of MMAC to traditional and multispectral images, and further shed light on large scale remote sensing images.

並列關鍵字

multi-level morphological active contour (MMAC) ； tree detection ； tree crowns delineation ； spectral analysis ； classification

參考文獻

[1]Lin C., Thomson G., Lo C.S., Yang M.S., “A multi-level morphological active contour algorithm for delineating tree crowns in mountainous forest,” Photogrammetric Engineering and Remote Sensing 77(3): 241-249, 2011.

[2]Chinsu Lin; Chein-Shun Lo; Thomson, G., "A textural modification of the MMAC algorithm for individual tree delineation in forest stand using aerial bitmap images," Image and Signal Processing (CISP), 2011 4th International Congress on , vol.3, no., pp.1604,1608, 15-17 Oct. 2011

[3]Rafael C. Gonzalez, Richard E. Woods, "Digital Image Processing," Upper Saddle River, NJ : Pearson/Prentice Hall, 2008.

[4]Kass, M., A. Witkin, and D. Terzopoilos, 1988. Snakes: Active contour models, International Journal of Computer Vision,1(4):321–331.

[5]Ganesh Saiprasad, "Spleen Segmentation and volume estimation using a gradient vector flow (GVF) based snake model, " ProQuest, UMI Dissertation Publishing, 2012.