6B), iron objects that litter the site are deposited by hundreds of visiting tourists each year. Small idiosyncrasies in this surface from animal burrows, a few looters’ holes, and unfilled early-20th-century archaeological excavations are reduced through repeated application of an adaptive box filter (replacing deviant values with local means) and application of a narrow-radius (0.5 m) low-pass filter. Surface depressions (houses, ditches, and bastions) thereby become negative in value with positive results representing raised berms that once formed walls surrounding houses or excavated earth from the defensive ditch, all centered about zero-valued “neutral” ground ( Fig. Huff exhibits a uniform slope that drops about 4.5 m across the village's width, which is eliminated through a 7.5 m radius high-pass filter that “flattens” the landscape. Menzer, in Innovation in Near-Surface Geophysics, 2019 7.3.1 Preprocessing These three components are convolved with each other to obtain the net PSF, as explained in Chapter 3, which should be used to calculate the LPF component in Eq. The resampling function used to register the two images (see Chapter 7) The remaining lower-resolution sensor PSF components, including image motion, optics, etc. The GIFOV of the lower-resolution sensor, e.g., a 3 × 3 square PSF when merging 10m panchromatic and 30 m multispectral ALI images The net PSF consists of three components, all sampled at the smaller GSI, Essentially, using the net relative PSF between the lower- and higher-resolution images maintains the local radiance from the original to the reconstructed images. This intuitive relationship can be derived mathematically ( Schowengerdt and Filiberti, 1994). (8-10), it would seem reasonable to match the LPF to the effective PSF of the multispectral image relative to that of the higher-resolution panchromatic image. These filters cannot be arbitrary because they define the radiometric normalization between the two images. Low- and high-pass filters must be applied to the higher-resolution image to implement spatial domain fusion.
0 kommentar(er)
