Hybrid Constraints of Pure and Mixed Pixels for Soft-Then-Hard Super-Resolution Mapping With Multiple Shifted Images

Abstract

Multiple shifted images (MSIs) have been widely applied to many super-resolution mapping (SRM) approaches to improve the accuracy of fine-scale land-cover maps. Most SRM methods with MSIs involve two processes: subpixel sharpening and class allocation. Complementary information from the MSIs has been successfully adopted to produce soft attribute values of subpixels during the subpixel sharpening process. Such information, however, is not used in the second process of class allocation. In this paper, a new class-allocation algorithm, named "hybrid constraints of pure and mixed pixels" (HCPMP), is proposed to allocate land-cover classes to subpixels using MSIs. HCPMP first determines the classes of subpixels that overlap with the pure pixels of auxiliary images in MSIs, after which the remaining subpixels are classified using information derived from the mixed pixels of the base image in MSIs. An artificial image and two remote sensing images were used to evaluate the performance of the proposed HCPMP algorithm. The experimental results demonstrate that HCPMP successfully applied MSIs to produce SRM maps that are visually closer to the reference images and that have greater accuracy than five existing class-allocation algorithms. Especially, it can produce more accurate SRM maps for high-resolution land-cover classes than low-resolution cases. The algorithm takes slightly less runtime than class allocation using linear optimization techniques. Hence, HCPMP provides a valuable new solution for class allocation in SRM using auxiliary data from MSIs.Index Terms-Hybrid constraints, multiple shifted images (MSIs), remotely sensed imagery, super-resolution mapping (SRM).

Object-Based Superresolution Land-Cover Mapping From Remotely Sensed Imagery

“…(8) to maximize the soft class values of the subpixels within each object, subjected to the class proportional constraints of each object in Eq. (9). Note that only mixed objects perform this model whereas the pure object is directly assigned to the same land-cover class to its all subpixels for saving computation time.…”

“…The overall accuracy (OA) metric was employed to quantitatively assess the accuracy of classified maps. Note that the calculation of OA in the first two experiment was only for mixed objects to avoid the influence of pure objects as pure objects provided no useful information in evaluating SRM method [9,29,49]. Both object-based hard and soft classifiers are used the k-nearest neighbor approach when classifying spectral remote sensing images in the three experiments [50].…”

“…Fortunately, super resolution mapping (SRM) (also termed as subpixel mapping) technique is proposed as a promising solution to mixed pixels [4]. SRM fist disaggregates each coarse pixel in fractional images (i.e., the output of pixel-based soft classification) into fine subpixels and then determines where the subpixels of each land-cover class spatially distribute within a pixel [6][7][8][9][10][11][12]. Over the past decades, many SRM methods have been proposed.…”

Designing an Experiment to Investigate Subpixel Mapping as an Alternative Method to Obtain Land Use/Land Cover Maps

“…The APP is an index with which to evaluate the areal spatial pattern, because most existing SPM methods are suited to areal features [16,18,24,25,40]. The APP is equal to the area of features with areal pattern, divided by the total area:…”

“…The accuracy might decrease with increasing zoom factor [6,24,40,41], and therefore we are interested in establishing at which zoom factor the accuracy stabilizes.…”