ICTACT Journal on Image and Video Processing

Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Shadow Detection from Aerial Imagery with Morphological Preprocessing and Pixel Clustering Methods


Affiliations
1 Department of Electronics and Telecommunication Engineering, Cummins College of Engineering for Women, India
2 Signal and Image Processing Group, Space Applications Centre, India
     

   Subscribe/Renew Journal


Building extraction from aerial imagery facilitates many geo-specialized tasks like Urban Planning, Map Generation and Disaster Management. Well planned cities ensure good sanitation, lesser pollution and hence, a better standard of living for its citizens. This is essential for developing countries which face a major crisis of urban migration and space crunch, and where planned cities would be a move towards smart living. The objective of this work is to segment building footprints from aerial images. Traditional pixel clustering algorithms like K-means, Color Quantization (CQ) and Gaussian Mixture Model (GMM) are implemented with inclusion of preprocessing steps for improved performance. These techniques are compared based on performance and time taken. The number of clusters/components are selected on the basis of Silhouette Score and Akaike Information Criterion/ Bayesian Information Criterion (AIC/BIC). A commonly encountered problem in building segmentation is misclassification of pixels due to shadows. This challenge is dealt by masking shadows using morphological operations as a part of preprocessing.

Keywords

Shadow Detection, K-Means, Colour Quantization, Gaussian Mixture Model, AIC/BIC.
Subscription Login to verify subscription
User
Notifications
Font Size

  • SpaceNet on AWS, Available at https://spacenetchallenge.github.io/datasets/spacenetBuildings-V2summary.html, Accessed at 2020.

  • Rahman Farnoosh, Gholamhossein Yari and Behnam Zarpak, “Image Segmentation using Gaussian Mixture Models”, International Journal of Engineering Science, Vol. 19, No. 2, pp. 1-16, 2008.

  • German Valenzuela, M. Emre Celebi and Gerald Schaefer, “Colour Quantization using Coreset Sampling”, Proceedings of IEEE International Conference on Systems, Man and Cybernetics, pp. 1-12, 2018.

  • Timur Sattarov and Mircea-Serban Pavel, “On the Importance of Preprocessing and Initialization in K-Means”, Proceedings of IEEE International Conference on Data Mining and its Applications, pp. 1-12, 2015.

  • M. Emre Celebi, “Effective Initialization of K-means for Color Quantization”, Proceedings of IEEE International Conference on Image Processing, pp. 1649-1652, 2009.

  • Sudhir Singh and Nasib Singh Gill, “Analysis and Study of K-Means Clustering Algorithm”, International Journal of Engineering Research and Technology, Vol. 2, No. 7, pp. 2546-2551, 2013.

  • Chinki, Chandhok, Soni Chaturvedi and Aleefia Khurshid, “An Approach to Image Segmentation using K-means Clustering Algorithm”, International Journal of Artificial Intelligence and Neural Network, Vol. 2, No. 2, pp. 1-13, 2012.

  • J. Peng, D. Zhang and Y. Liu, “An Improved Snake Model for Building Detection from Urban Aerial Images”, Pattern Recognition Letters, Vol. 26, No. 5, pp. 587-595, 2005.

  • S. Ahmadi, M.V. Zoej, H. Ebadi, H.A. Moghaddam and A. Mohammadzadeh, “Automatic Urban Building Boundary Extraction from High Resolution Aerial Images using an Innovative Model of Active Contours”, International Journal of Applied Earth Observation and Geoinformation, Vol. 12, No. 3, pp. 150-157, 2010.

  • P. Kaiser, J.D. Wegner, A. Lucchi, M. Jaggi, T. Hofmann and K. Schindler, “Learning Aerial Image Segmentation from Online Maps”, IEEE Transactions on Geoscience and Remote Sensing, Vol. 55, No. 11, pp. 6054-6068, 2017.

  • Thiago Statella and Erivaldo Antônio da Silva, “Shadows and Clouds Detection in High Resolution Images using Mathematical Morphology”, Proceedings of International Annual Conference on Imaging and Geospatial Information Society, pp. 1-7, 2018.

  • Tapas Kanungo and Nathan S. Natanyahu, “An Efficient K-means Clustering Algorithm: Analysis and Implementation”, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No. 7, pp. 1-14, 2002.

  • Henryk Palus, “Color Image Quantization by the K-Means Algorithm”, in Workshop Farbbildverarbeitung, 7-8 October 2004.

  • M. Emre Celebi, “Improving the Performance of K-Means for Color Quantization”, Image and Vision Computing, Vol. 29, No. 4, pp. 260-271, 2011.

  • Liu Jun and Zhang Haili, “Image Segmentation using a Local GMM in a Variational Framework”, Journal of Mathematical Imaging and Vision, Vol. 46, No. 2, pp. 1-15, 2013.


Abstract Views: 190

PDF Views: 0




  • Shadow Detection from Aerial Imagery with Morphological Preprocessing and Pixel Clustering Methods

Abstract Views: 190  |  PDF Views: 0

Authors

Ashwini M. Deshpande
Department of Electronics and Telecommunication Engineering, Cummins College of Engineering for Women, India
Manali Gaikwad
Department of Electronics and Telecommunication Engineering, Cummins College of Engineering for Women, India
Saudamini Patki
Department of Electronics and Telecommunication Engineering, Cummins College of Engineering for Women, India
Aditee Rathi
Department of Electronics and Telecommunication Engineering, Cummins College of Engineering for Women, India
Sampa Roy
Signal and Image Processing Group, Space Applications Centre, India

Abstract


Building extraction from aerial imagery facilitates many geo-specialized tasks like Urban Planning, Map Generation and Disaster Management. Well planned cities ensure good sanitation, lesser pollution and hence, a better standard of living for its citizens. This is essential for developing countries which face a major crisis of urban migration and space crunch, and where planned cities would be a move towards smart living. The objective of this work is to segment building footprints from aerial images. Traditional pixel clustering algorithms like K-means, Color Quantization (CQ) and Gaussian Mixture Model (GMM) are implemented with inclusion of preprocessing steps for improved performance. These techniques are compared based on performance and time taken. The number of clusters/components are selected on the basis of Silhouette Score and Akaike Information Criterion/ Bayesian Information Criterion (AIC/BIC). A commonly encountered problem in building segmentation is misclassification of pixels due to shadows. This challenge is dealt by masking shadows using morphological operations as a part of preprocessing.

Keywords


Shadow Detection, K-Means, Colour Quantization, Gaussian Mixture Model, AIC/BIC.

References