PySizer: resizing raster in eigenspace using Python

A common practice in spatial analysis, data fusing, and geophysical interpretation is the need to handle raster datasets of different spatial resolution. Often, raster data are resampled through interpolation for better spatial representation and feature identification at both local and regional scales, but interpolation can result in artifact such as fewer blocks or mosaics, boundary blurring, edge halos, rings, and signal aliasing. In this study, we present the PySizer program developed using the Python programming language for data interpolation, with advanced performance. First, it has high interpolation accuracy by using the inverse spatial principal analysis (isPCA) in the eigenspace. Secondly, it has high calculation performance by using functions inside scientific Python modules, such as matrix operations of the NumPy. Thirdly, it supports most of currently available raster data formats and can manage spatial references information. PySizer is applicable to many fields, such as geophysical or other scientific data processing, remote sensing, (internet) video stream interpolation, and high quality image printing, to name a few. The tool was tested in this study using a potential-field data and bathymetric data, and the results were evaluated through visual inspection and statistical analysis, demonstrating high performance in accuracy and efficiency in resizing raster datasets. The PySizer source codes are freely available from public website or by contacting the authors for the latest version.