Numerical recipes in C (2nd ed.): the art of scientific computing
Numerical recipes in C (2nd ed.): the art of scientific computing
Digital image processing
Remote Sensing, Third Edition: Models and Methods for Image Processing
Remote Sensing, Third Edition: Models and Methods for Image Processing
| Hi-index | 0.05 |
It has been shown that spatial subpixel analysis can be used to enhance images of fine structured landscapes. This method is based on the geometric description of object boundaries that intersect pixels and thus lead to mixed pixels. The parameters of the geometric model describing the underlying scene are estimated by means of an optimization algorithm. The applicability of the method depends on the relationship between the size of the remotely sensed objects and the pixel size. Possible applications include pre-processing for classification to reduce the percentage of mixed pixels, vector segmentation approaches and image fusion techniques. The distribution of grey values in the neighbourhood of any mixed pixel not only depends on the parameters of the geometric model of land cover boundaries, but also on the spatial response of the sensor. Therefore, knowledge about the sensor point spread function can be used to enhance performance of spatial subpixel analysis. If the parameters of the point spread function are included as unknowns in the fitting problem, the optimization may give an estimate of the spatial response. For this purpose we assume a Gaussian-shaped point spread function with two parameters, namely the standard deviations along the two image axes, and search for an optimal solution for fitting the model parameters to the actual scene. In this contribution we describe a method for estimating the point spread function by spatial subpixel analysis. We apply the algorithm to different synthetic and real images. The sensitivity of the method to varying input patterns is discussed and the improvement of the results of spatial subpixel analysis by taking into account the point spread function determined in the described way is illustrated.