Teerasit Kasetkasem

Channel aware decision fusion in wireless sensor networks

Information fusion by utilizing multiple distributed sensors is studied in this work. Extending the classical parallel fusion structure by incorporating the fading channel layer that is omnipresent in wireless sensor networks, we derive the likelihood ratio based fusion rule given fixed local decision devices. This optimum fusion rule, however, requires perfect knowledge of the local decision performance indices as well as the fading channel. To address this issue, two alternative fusion schemes, namely, the maximum ratio combining statistic and a two-stage approach using the Chair-Varshney fusion rule, are proposed that alleviate these requirements and are shown to be the low and high signal-to-noise ratio (SNR) equivalents of the likelihood-based fusion rule. To further robustify the fusion rule and motivated by the maximum ratio combining statistics, we also propose a statistic analogous to an equal gain combiner that requires minimum a priori information. Performance evaluation is performed both analytically and through simulation.

An image change detection algorithm based on Markov random field models

This paper addresses the problem of image change detection (ICD) based on Markov random field (MRF) models. MRF has long been recognized as an accurate model to describe a variety of image characteristics. Here, we use the MRF to model both noiseless images obtained from the actual scene and change images (CIs), the sites of which indicate changes between a pair of observed images. The optimum ICD algorithm under the maximum a posteriori (MAP) criterion is developed under this model. Examples are presented for illustration and performance evaluation.

Fusion of decisions transmitted over fading channels in wireless sensor networks

Information fusion by utilizing multiple distributed sensors is studied. We derive the optimal likelihood based fusion statistic for a parallel decision fusion problem with fading channel assumption. This optimum fusion rule, however, requires perfect knowledge of the local decision performance indices as well as the fading channel. Several alternatives are presented that alleviate these requirements. At low SNR, the likelihood based fusion rule reduces to a form analogous to a maximum ratio combining statistic; while at high SNR, it leads to a two-stage approach using the well known Chair-Varshney fusion rule. A third alternative, in the form of an equal gain combiner, is also proposed; it requires the least amount of information regarding the sensor/channel. Simulation shows that the two-stage approach, which considers the communication and decision fusion as two independent stages, suffers performance loss compared with the other two alternatives for a practical SNR range.

Estimation of Fuzzy Error Matrix Accuracy Measures Under Stratified Random Sampling

A fuzzy error matrix can be used to summarize accuracy assessment information when both the map and reference data are labeled using a soft classification. Accuracy measures analogous to the familiar overall, users, and producers accuracies of a hard classification can be derived from a fuzzy error matrix. The formulas for estimating the fuzzy error matrix and accompanying accuracy measures depend on the sampling design used to collect the reference data. In this paper, these estimation formulas are derived for a stratified random sampling, a design commonly implemented in practice. A simulation study is conducted to confirm the validity of the stratified sampling estimators.

Automatic Rice Crop Height Measurement Using a Field Server and Digital Image Processing

Rice crop height is an important agronomic trait linked to plant type and yield potential. This research developed an automatic image processing technique to detect rice crop height based on images taken by a digital camera attached to a field server. The camera acquires rice paddy images daily at a consistent time of day. The images include the rice plants and a marker bar used to provide a height reference. The rice crop height can be indirectly measured from the images by measuring the height of the marker bar compared to the height of the initial marker bar. Four digital image processing steps are employed to automatically measure the rice crop height: band selection, filtering, thresholding, and height measurement. Band selection is used to remove redundant features. Filtering extracts significant features of the marker bar. The thresholding method is applied to separate objects and boundaries of the marker bar versus other areas. The marker bar is detected and compared with the initial marker bar to measure the rice crop height. Our experiment used a field server with a digital camera to continuously monitor a rice field located in Suphanburi Province, Thailand. The experimental results show that the proposed method measures rice crop height effectively, with no human intervention required.

An Optimum Land Cover Mapping Algorithm in the Presence of Shadows

Occurrence of shadowy pixels in remote sensing images is a common phenomenon particularly with passive sensors. In these cases, analysts may mistakenly treat these pixels as a separate land cover class. This may result in the loss of information present in the shadow pixels A better approach may be to correct light intensity values in shadowy pixels and use the light-corrected image to produce the land cover map. Most light intensity correction algorithms are not designed to optimize classification performance. Consequently, the accuracy of the resulting land cover map may be degraded. To address this problem, this paper proposes an algorithm that employs the maximum a posteriori criterion for classifying a multispectral image in the presence of shadows. The observed image is assumed to be the product of a shadow-free image with a light intensity image along with an additive measurement noise. The main purpose of this algorithm is to find the most likely land cover map along with the shadow-free image and light intensity image as byproducts. Our results show that a large number of misclassified pixels can be corrected. Furthermore, in the shadow-free image, the materials in the shadowy regions can also be successfully reconstructed.

Statistical characterization of clutter scenes based on a Markov random field model

The problem of clutter region identification based on Markov random field (MRF) models is addressed. Observations inside each clutter region are assumed homogenous, i.e., observations follow a single probability distribution. Our goal is to partition clutter scenes into homogenous regions and to determine this underlying probability distribution. Metropolis-Hasting and reversible jump Markov chain (RJMC) algorithms are used to search for solutions based on the maximum a posteriori (MAP) criterion. Several examples illustrate the performance of our algorithm.

Sub-pixel land cover mapping based on Markov random field models

Occurrence of mixed pixels in remote sensing images is a common phenomenon particularly in coarse spatial resolution images. In these cases, sub-pixel or soft classification may be preferred over conventional hard classification. However, sub-pixel classification fails to account for the spatial distribution of class proportions within the pixel. A better approach may be to generate a land cover map at a finer resolution from the coarse resolution images based on image models that accurately characterize the spatial distribution of the classes. The resulting fine resolution map may be called a sub-pixel or super resolution map. In this paper, an approach based on Markov random fields is introduced to generate sub-pixel land cover maps from remote sensing images dominated by mixed pixels.

Fusion and registration of THEOS multispectral and panchromatic images

This article presents a new method for the fusion and registration of THEOS (Thailand Earth Observation Satellite) multispectral and panchromatic images in a single step. In the usual procedure, fusion is an independent process separated from the registration process. However, both image registration and fusion can be formulated as estimation problems. Hence, the registration parameters can be automatically tuned so that both fusion and registration can be optimized simultaneously. Here, we concentrate on the relationship between low-resolution multispectral and high-resolution panchromatic imagery. The proposed technique is based on a statistical framework. It employs the maximum a posteriori (MAP) criterion to jointly solve the fusion and registration problem. Here, the MAP criterion selects the most likely fine resolution multispectral and mapping parameter based on observed coarse resolution multispectral and fine resolution panchromatic images. The Metropolis algorithm was employed as the optimization algorithm to jointly determine the optimum fine resolution multispectral image and mapping parameters. In this work, a closed-form solution that can find the fused multispectral image with correcting registration is also derived. In our experiment, a THEOS multispectral image with high spectral resolution and a THEOS panchromatic image with high spatial resolution are combined to produce a multispectral image with high spectral and spatial resolution. The results of our experiment show that the quality of fused images derived directly from misaligned image pairs without registration error correction can be very poor (blurred and containing few sharp edges). However, with the ability to jointly fuse and register an image pair, the quality of the resulting fused images derived from our proposed algorithm is significantly improved, and, in the simulated cases, the fused images are very similar to the original high resolution multispectral images, regardless of the initial registration errors.

A Joint Land Cover Mapping and Image Registration Algorithm Based on a Markov Random Field Model

Traditionally, image registration of multi-modal and multi-temporal images is performed satisfactorily before land cover mapping. However, since multi-modal and multi-temporal images are likely to be obtained from different satellite platforms and/or acquired at different times, perfect alignment is very difficult to achieve. As a result, a proper land cover mapping algorithm must be able to correct registration errors as well as perform an accurate classification. In this paper, we propose a joint classification and registration technique based on a Markov random field (MRF) model to simultaneously align two or more images and obtain a land cover map (LCM) of the scene. The expectation maximization (EM) algorithm is employed to solve the joint image classification and registration problem by iteratively estimating the map parameters and approximate posterior probabilities. Then, the maximum a posteriori (MAP) criterion is used to produce an optimum land cover map. We conducted experiments on a set of four simulated images and one pair of remotely sensed images to investigate the effectiveness and robustness of the proposed algorithm. Our results show that, with proper selection of a critical MRF parameter, the resulting LCMs derived from an unregistered image pair can achieve an accuracy that is as high as when images are perfectly aligned. Furthermore, the registration error can be greatly reduced.