Super-resolution target identification from remotely sensed images using a Hopfield neural network

Abstract

Abstract-Fuzzy classification techniques have been developed recently to estimate the class composition of image pixels, but their output provides no indication of how these classes are distributed spatially within the instantaneous field of view represented by the pixel. As such, while the accuracy of land cover target identification has been improved using fuzzy classification, it remains for robust techniques that provide better spatial representation of land cover to be developed. Such techniques could provide more accurate land cover metrics for determining social or environmental policy, for example. The use of a Hopfield neural network to map the spatial distribution of classes more reliably using prior information of pixel composition determined from fuzzy classification was investigated. An approach was adopted that used the output from a fuzzy classification to constrain a Hopfield neural network formulated as an energy minimization tool. The network converges to a minimum of an energy function, defined as a goal and several constraints. Extracting the spatial distribution of target class components within each pixel was, therefore, formulated as a constraint satisfaction problem with an optimal solution determined by the minimum of the energy function. This energy minimum represents a "best guess" map of the spatial distribution of class components in each pixel. The technique was applied to both synthetic and simulated Landsat TM imagery, and the resultant maps provided an accurate and improved representation of the land covers studied, with root mean square errors (RMSEs) for Landsat imagery of the order of 0.09 pixels in the new fine resolution image recorded. As such, we show how, by using a Hopfield neural network, more accurate measures of land cover targets can be obtained compared with those determined using the proportion images alone. The Hopfield neural network used in this way represents a simple, robust, and efficient technique, and results suggest that it is a useful tool for identifying land cover targets from remotely sensed imagery at the subpixel scale.

Super-resolution land cover pattern prediction using a Hopfield neural network

“…This allows performance comparison with the class proportion-defined initialisation and does not introduce any possible unnecessary bias into the result, which may occur using (i) should estimated class proportions be inaccurate. Tatem et al (2001) demonstrated no significant differences in results from the two initialisation techniques.…”

“…As the pattern prediction technique represents an extension of the method described in Tatem et al (2001), results of both techniques were compared to highlight the need for a dual approach. To determine the success of each technique at recreating the spatial pattern exhibited in the target images, plots of lag against semivariance (a variogram; Curran & Atkinson, 1998), calculated using Eq.…”

“…Consequently, while these features can be detected within a pixel by soft classification techniques, surrounding pixels hold no information for inference of spatial relationships to aid their mapping. Therefore, the technique presented in this paper attempts to overcome this problem and to present a novel and effective solution to super-resolution land cover pattern prediction from remotely sensed images, as well as an extension to the technique proposed by Tatem et al (2001). This method is based on prior information on the spatial arrangement of land cover.…”

“…Results showed a certain degree of success, but problems were noted where land cover features were smaller than a pixel. Tatem et al (2001) presented a technique that allowed prior information to be included in the prediction process. A Hopfield neural network was formulated as an energy minimisation tool to predict the land cover distribution within each pixel.…”

“…However, there exist practical limitations (Tatem et al, 2001). Perhaps the biggest drawback of obtaining land cover information from remotely sensed images relates to scale.…”