Performance Comparison of Object- and Pixel-based Land CoverClassifications using a Color Infrared Aerial Photograph

다국어 초록 (Multilingual Abstract)

This study introduces an object-based image analysis (OBIA) to perform land cover classifications with abenchmark of a pixel-based approach. The classification performance of the two approaches is evaluated with overallaccuracy, individual accuracy and a statistical z test. In addition, the research investigates the effect of segmentationscale on the quality of segmentation and the results of classifications. The study discovered that the OBIAoutperformed the pixel-based approach with an increase of 11.4 % in overall accuracy (a Kappa of 0.13). In general, theproducer’s and user’s accuracies of the OBIA were also improved with a range of 6.1 % to 41.8 %. A pair-wise z testrevealed that the two classification results were significantly different at a confidence level of 99 %, with a z statistic of3.14. The quality of segmentation was found to directly affect the results of object-based classifications. Overall, theOBIA will be a potential methodology to improve land cover classifications from a very high spatial resolution (VHR)remote sensing imagery, and this approach can closely couple remote sensing with geographic information systems.

참고문헌 (Reference)

1 Definiens, "eCognition User Guide 4" Definiens AG 2004

2 Blaschke,T., "What’s wrong with pixels? some recent developments interfacing remote sensing and GIS" 6 : 12-17, 2001

3 Franklin, S. E., "Using spatial co-occurrence texture to increase forest structure and species composition classification accuracy" 67 : 849-855, 2001

4 Fisher,P., "The pixel: a snare and a delusion" 18 (18): 679-685, 1997

5 Ferro, C. J. S., "Scale and texture in digital image classification" 68 (68): 51-63, 2002

6 Lillesand,T.M., "Remote sensing and image interpretation. 4th edition" John Wiley & Sons 2000

7 Kim, M., "Objectbased vegetation type mapping from an orthorectified multispectral IKONOS image using ancillary information" Commission IV, WG IV/4 on Proceeding of GEOBIA 2008 - Pixels, Objects, Intelligence: GEOgrpahic Objec 2008b

8 Yu, Q., "Object-based detailed vegetation classification with airborne high spatial resolution remote sensing imagery" 72 (72): 799-811, 2006

9 Blaschke,T., "Object-based contextual image classification built on image segmentation" Proceedings of the 2003 IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data 113-119, 2003

10 Benz, U. C., "Multi-resolution, object-oriented fuzzy analysis of remote sensing data for GISready information" 58 : 239-258, 2004

11 Kimes, D. S., "Mapping secondary tropical forest and forest age from SPOT HRV data" 20 (20): 3625-3640, 1999

12 Woodcock, C. E., "Mapping forest vegetation using Landsat TM imagery and a canopy reflectance model" Remote Sensing 1994

13 Wang, L., "Integration of object-based and pixel-based classification for mapping mangroves with IKONOS imagery" 25 (25): 5655-5668, 2004

14 Dorren, L. K. A., "Improved Landsat-based forest mapping in steep mountainous terrain using object-based classification" 183 : 31-46, 2003

15 Swain,P.H., "Fundamentals of pattern recognition in remote sensing in : Remote sensing: the quantitative approach" McGraw-Hill 1978

16 Kim, M., "Forest type mapping using object-specific texture measures from multispectral IKONOS imagery: segmentation quality and image classification issues" Photogrammetric Engineering & Remote Sensing In print

17 Kim, M., "Estimation of optimal image object size for the segmentation of forest stands with multispectral IKONOS imagery" Object-Based Image Analysis - Spatial concepts for 2008a

18 Kim, M., "Determination of optimal scale parameter for alliance-level forest classification of multispectral IKONOS image. Commission IV, WG IV/4 on Proceeding of 1st OBIA Conference" International Soci 2006

19 Ryherd, S., "Combining spectral and texture data in the segmentation of remotely sensed images" 62 (62): 181-194, 1996

20 Connors,K.F., "Classification of geomorphic features and landscape stability in Northwestern New Mexico using simulated SPOT imagery" 22 (22): 187-207, 1987

21 Townshend, J. R. G., "Beware of per-pixel characterization of land cover" 21 (21): 839-843, 2000

22 Carleer,A.P., "Assessment of very high spatial resolution satellite image segmentations" 71 (71): 1285-1294, 2005

23 Congalton, R. G., "Assessing the accuracy of remotely sensed data: principles and practices" Lewis Publishers 1999

24 Hay, G. J., "An object-specific image-texture analysis of H-resolution forest imagery" 55 : 108-122, 1996

25 Burnett,C., "A multi-scale segmentation/object relationship modeling methodology for landscape analysis" 168 : 233-249, 2003

26 Feitosa, C. U., "A genetic approach for the automatic adaptation of segmentation parameters. Commission IV, WG IV/4 on Proceeding of 1st OBIA Conference" International Society for 2006

27 Meinel, G., "A comparison of segmentation programs for high resolution remote sensing data. Commission VI in Proceeding of XXth ISPRS Congress" International Society for Photogrammetry and Remote Sens 2004