Comparing Pan-Sharpening Algorithms to Access an Agriculture Area: A Mississippi Case Study

Abstract

Numerous satellites collect imagery of the Earth's surface daily, providing information to the public and private sectors. The fusion (pan-sharpening) of high-resolution panchromatic satellite imagery with lower-resolution multispectral satellite imagery has shown promise for monitoring natural resources and farming areas. It results in new imagery with more detail than the original multispectral or panchromatic images. In agricultural areas in Mississippi, landscapes can range from complex mixtures of vegetation and built-up areas to dense vegetative regions. More information is needed on pan-sharpened imagery for assessing landscapes in rural areas of Mississippi. WorldView 3 satellite imagery consisting of landscapes commonly found in rural areas of Mississippi was subjected to 17 pan-sharpening algorithms. The pan-sharpened images were compared qualitatively and quantitatively with three quality indices: 1) Erreur Relative Globale Addimensionelle de Synthese; 2) Universal Image Quality Index; 3) Bias. À trous wavelet transform with the injection model 3 and hyperspherical color spaced fusion methods were ranked among the best for maintaining image integrity for qualitative and quantitative analyses. The optimized high-pass filter method was often ranked last by the quality indices. The smoothing filter-based intensity modulation algorithm and the gaussian modulation transfer function match filtered with high-pass modulation injection model added artifacts to the images. Pansharpened satellite imagery has great potential to enhance the survey of Mississippi's agricultural areas. The key to success is selecting an image fusion process that increases spatial content while not compromising the image integrity.

Selected Agricultural Analyses Based on Data from MultiSen-PL, the Multi-sensor Airborne Remote Sensing Station

“…It demonstrates its universal nature and suitability for sharpening images captured by various sensors recording in bands outside the visible range. Notably, Fletcher (2023), utilising WordView 3 farmland images, demonstrated that applying this method to convert 1.2 m spectral channels to 0.3 m resolution yielded some of the best results among the 17 pansharpening methods tested. The slight blurring observed in satellite images sharpened using the HCS technique diminishes with increasing geometric resolution.…”

Measurement Accuracy and Improvement of Thematic Information from Unmanned Aerial System Sensor Products in Cultural Heritage Applications

“…In this case we are dealing with a low-quality fused image, which differs significantly from the MS orthophotomosaic. All the above limits may, as mentioned above, be modified but in any case, the index values should be less than three in order for a fused image to be characterized in terms of its quality and/or used for classification [63,[71][72][73][74][75][76][77][78][79][80][81][82][83][84].…”