DOI QR코드

DOI QR Code

Satellite Image Classification Based on Color and Texture Feature Vectors

칼라 및 질감 속성 벡터를 이용한 위성영상의 분류

  • Published : 1999.09.01

Abstract

The Brightness, color and texture included in a multispectral satellite data are used as important factors to analyze and to apply the image data for a proper use. One of the most significant process in the satellite data analysis using texture or color information is to extract features effectively expressing the information of original image. It was described in this paper that six features were introduced to extract useful features from the analysis of the satellite data, and also a classification network using the back-propagation neural network was constructed to evaluate the classification ability of each vector feature in SPOT imagery. The vector features were adopted from the training set selection for the interesting region, and applied to the classification process. The classification results showed that each vector feature contained many merits and demerits depending on each vector's characteristics, and each vector had compatible classification ability. Therefore, it is expected that the color and texture features are effectively used not only in the classification process of satellite imagery, but in various image classification and application fields.

위성에서 관측된 다중분광 위성영상 데이터를 이용목적에 따라 분석하고 활용하기 위해서는 영상 자체에 내포된 밝기, 칼라, 질감 등 다양한 특징들이 중요한 정보원으로 이용되고 있다. 특히 질감이나 칼라정보를 이용한 위성영상의 분석과정에서 가장 중요한 문제는 원 영상의 정보를 효율적으로 표현하는 속성을 추출하여 적절히 활용하는 것이다. 따라서 본 논문에서는 위성영상 분석에 유용하게 사용할 수 있는 6개의 속성 벡터들을 선정한 다음 SPOT 위성에서 관측된 영상을 이용하여 각각의 속성들에 대한 분별력을 평가하기 위하여 역전파 신경망(Back-propagation Neural Network)을 이용한 분류 네트워크를 구성하였고, 실험하고자 하는 지역에 대한 훈련집합 선택시 선정된 여섯 개이 속성 벡터들을 분류에 사용될 특징으로 선택하였다. 분류 실험을 수행한 결과 각각의 벡터 속성들은 개개의 특성에 따라 많은 장단을 내포하고 있었으며, 전반적으로는 비교적 정확한 분류결과를 나타내었다. 따라서 칼라 및 질감 속성 벡터들은 위성영상의 분류과정에 효과적으로 사용될 수 있음은 물론 다양한 영상분석 및 응용분야에서도 유용하게 이용될 수 있을 것으로 기대된다.

Keywords

References

  1. Int. J. of Remote Sensing v.16 no.9 The effect of training set size and composition on artificial neural network classification G. M. Foody;M. B. Mcculloch;W. B. Yates
  2. Remote Sensing digital Image Analysis John A. Richards
  3. Proceedings of SPIE, Bejing, China, Electronic Imaging and Multimedia System Ⅱ, 3561 Vector feature for Textured Color Images and Their Application to Content-Based Image Retrievals J. Kim;S. Kim;W. You;J. Lee
  4. IEEE Int. Conf. on CVPR Image Indexing Using Color Correlograms J. Huang;S. Ravi Kumar;Mandar Mitra;Wei-Jing Zhu;Ramin Zabih
  5. IEEE Trans. on Geoscience and Remote sensing v.28 no.5 A Neural Network Approach to Cloud classification Jonathan Lee;Ronald C. Weger;Sailes K. Sengupta;Ronald M. Welch
  6. IEEE Int. Conf. on Image Processing Local Color and Texture Extraction and Spatial Query J. R. Smith;S. F. Chang
  7. Symposium on Electric Imaging: Science and Technology-Storage & retrieval for Image and Video Database Ⅳ. IS&T/SPIE Proceedings 2670. Tools and techniques for color images Retrieval J. R. Smith;S. F. Chang
  8. Proc. of the Workshop on Applications of Computer Vision Color and Texture Fusion: Application to Aerial Image Segmentation and GIS Updating M. P. Dubuisson-Jolly;A. Gupta
  9. Computer processing of remotely sensed images Prul M. Marther
  10. Pattern Recognition v.25 no.8 Performance Evaluation for Four Classes of Textural Features P. P. Ohanian;R. C. Dubes
  11. IEEE Trans. on Systems, Man and Cybemetics v.SMC-3 no.6 Textural Features for Image Classification Robert M. Haralick;K. Shanmugam;Its hak Dinstein
  12. Proceedings of IEEE v.67 no.5 Statistical and Structural Approaches to Texture Robert M. Haralick
  13. PE & RS v.49 Classification of Arid Geomorphic Surfaces Using Landsat Spectral and Textural Features Shin E. H.;R. A. Schowengerdt