ABSTRACT
This chapter examines high-pass filtering for edge and contour detection, techniques that are used to implement low-level feature matching for example, image co-registration. It shows how to access the highly optimized image processing algorithms of the Open Source Computer Vision Library from the Python interpreter and look at similar functionality on the Google Earth Engine Python application programing interface. Edge detection is often used in conjunction with feature extraction for scene analysis and for image segmentation. Many common satellite platforms supply co-registered panchromatic images with considerably higher ground resolution than that of the multispectral bands. Conventional image fusion techniques, such as the well-known HSV-transformation, can be used to sharpen the spectral components. Panchromatic sharpening using principal components analysis (PCA) is similar to the HSV method. Image sharpening using HSV, Brovey, PCA and the closely related Gram–Schmidt transformation are available in most commercial image processing environments such as ENVI or Geomatica.
