The use of texture for image classification of black & white air photographs

Abstract

The use of black & white air photographs for the production of historic land cover maps can be done by image classification, using additional texture features. In this paper we evaluate the importance of a number of parameters in the image classification process based on texture, such as the quantization level, the window size used to produce the texture features, the feature selection criteria and the image spatial resolution. The evaluation was performed using 4 photographs from the 1950s. The influence of the classification method, the number of classes searched for in the images and the post-processing tasks were also investigated. The importance of each of these parameters for the classification accuracy was evaluated by cross validation. The selection of the best parameters was performed based on the validation results, and also on the computation load involved for each case and the end user requirements. An average accuracy of 85.2% was achieved for 4 land cover classes.New Developments and Challenges in Remote Sensing, Z. Bochenek (ed.) ß2007 Millpress, Rotterdam,

Automated detection and mapping of avalanche deposits using airborne optical remote sensing data

“…A detailed overview of this method is given by (Haralick et al, 1973;Haralick, 1979;Tuceryan and Jain, 1998). This approach has shown good results for the classification of urban structures (Karathanassi et al, 2000;Pesaresi, 2000), agricultural structures (Caridade et al, 2008;Delenne et al, 2008) and sea ice (Soh and Tsatsoulis, 1999). Local texture variations are helpful indicators for rough snow surfaces.…”

Discriminating small wooded elements in rural landscape from aerial photography: a hybrid pixel/object-based analysis approach

“…The addition of texture has become quite common for land-cover mapping (Marceau et al 1990, Puissant et al 2005. It has been also widely used in forestry and ecological mapping applications (St Onge and Cavayas 1995, Franklin et al 2000, Coburn and Roberts 2004, especially when digital aerial photographs are required (Wulder et al 1998, Caridade et al 2007. Generally, the resolution of aerial photographs is finer than or similar to the size of the trees in the image.…”

Seasonal Land Cover Dynamics in Beijing Derived from Landsat 8 Data Using a Spatio-Temporal Contextual Approach

“…NDVI is calculated as: NDVI = (NIR − RED) / (NIR + RED) [21], where NIR is band 5 of Landsat 8, and RED is band 4. Three texture features, mean, contrast and entropy of grey-level co-occurrence matrix (GLCM) [22][23][24], were calculated from bands 5 and 8, respectively. Parameters for co-occurrence matrix were set as follows: grey levels = 16, window size = 7, and offset = (1, 1).…”