Xiaolong Dai

The effects of image misregistration on the accuracy of remotely sensed change detection

Image misregistration has become one of the significant bottlenecks for improving the accuracy of multisource data analysis, such as data fusion and change detection. In this paper, the effects of misregistration on the accuracy of remotely sensed change detection were systematically investigated and quantitatively evaluated. This simulation research focused on two interconnected components. In the first component, the statistical properties of the multispectral difference images were evaluated using semivariograms when multitemporal images were progressively misregistered against themselves and each other to investigate the band, temporal, and spatial frequency sensitivities of change detection to image misregistration. In the second component, the ellipsoidal change detection technique, based on the Mahalanobis distance of multispectral difference images, was proposed and used to progressively detect the land cover transitions at each misregistration stage for each pair of multitemporal images. The impact of misregistration on change detection was then evaluated in terms of the accuracy of change detection using the output from the ellipsoidal change detector. The experimental results using Landsat Thematic Mapper (TM) imagery are presented. It is interesting to notice that, among the seven TM bands, band 4 (near-infrared channel) is the most sensitive to misregistration when change detection is concerned. The results from false change analysis indicate a substantial degradation in the accuracy of remotely sensed change detection due to misregistration. It is shown that a registration accuracy of less than one-fifth of a pixel is required to achieve a change detection error of less than 10%.

A feature-based image registration algorithm using improved chain-code representation combined with invariant moments

A new feature-based approach to automated image-to-image registration is presented. The characteristic of this approach is that it combines an invariant-moment shape descriptor with improved chain-code matching to establish correspondences between the potentially matched regions detected from the two images. It is robust in that it overcomes the difficulties of control-point correspondence by matching the images both in the feature space, using the principle of minimum distance classifier (based on the combined criteria), and sequentially in the image space, using the rule of root mean-square error (RMSE). In image segmentation, the performance of the Laplacian of Gaussian operators is improved by introducing a new algorithm called thin and robust zero crossing. After the detected edge points are refined and sorted, regions are defined. Region correspondences are then performed by an image-matching algorithm developed in this research. The centers of gravity are then extracted from the matched regions and are used as control points. Transformation parameters are estimated based on the final matched control-point pairs. The algorithm proposed is automated, robust, and of significant value in an operational context. Experimental results using multitemporal Landsat TM imagery are presented.

Development of a feature-based approach to automated image registration for multitemporal and multisensor remotely sensed imagery

A new feature-based approach to automated multitemporal and multisensor image registration is presented. The characteristics of this technique is that it combines moment invariant shape descriptors with modified chain code correlation to establish the correspondences between potential matched regions in two images. It also overcomes the difficulties in control point correspondence in image matching caused by the problem of feature inconsistency. In image segmentation, the authors use the improved Laplacian of Gaussian (LoG) zero-crossing edge detector. Feature matching is done in both feature space and image space based on moment invariant distance and improved chain code correlation. The centers of gravity are then extracted from matched regions and used as control points. The final transformation parameters are estimated based on the final matched control points. Experimental results using multitemporal Landsat TM imagery are presented.

Development of a new automated land cover change detection system from remotely sensed imagery based on artificial neural networks

The research is designed to develop and implement the algorithms for an automated spatial change information extraction system from remotely sensed imagery based on artificial neural networks. First, the authors investigate the suitability of the application of neural networks in automated change detection using TM imagery and its related network design problems unique to change detection. They then develop a neural networks-based change detection system using backpropagation training algorithm. This trained network is then able to efficiently detect land cover changes and provide complete information about the nature of change. Based on their experiments, it has been proven that this technique is successful and has immense implications on land cover change detection and quantification at all levels of applications ranging from local to global in scale.

Quantification of the impact of misregistration on the accuracy of remotely sensed change detection

The impact of misregistration on the accuracy of change detection is quantitatively investigated using TM imagery. This simulation study focuses on two interconnected issues. First, the statistical properties of the difference images are evaluated using semivariograms when images are progressively misregistered in order to investigate the band sensitivity, temporal sensitivity, and spatial frequency sensitivity of change detection to misregistration. The ellipsoidal change detection technique is then proposed and used to progressively detect the land cover transitions at each misregistration stage for each image. The impact of misregistration on change detection is then evaluated in terms of the accuracy of change detection using the output from the ellipsoidal change detector.

Techniques for automated extraction of roadway inventory features from high-resolution satellite imagery

Abstract The emergence of high‐resolution satellite imagery is attracting new applications which can take advantage of remotely sensed data for mapping, inventory, and change detection. Automated collection of roadway inventory features is one such application. To this end, it is important to investigate the performance of conventional feature extraction techniques when applied to high‐resolution images and to develop new techniques for extraction of roadway features using one‐meter, or higher, resolution imagery. In this paper, classification‐ based and edge detection‐based techniques potential for automated extraction of roadway features from high‐resolution satellite imagery are described, tested, and evaluated. Of possible techniques, the applicability of conventional classification algorithms, the Thin and Robust Zero‐Crossing edge detector based on the Laplacian of Gaussian operator, and seeded region growing segmentation is investigated. The advantages and disadvantages of each technique for extrac...

Requirements and techniques for an automated change detection system

To enhance the ability of remote sensing system to provide accurate, timely, and complete geospatial information at regional and/or global scale, an automated change detection system has been and will continue to be one of the important yet challenging problems in remote sensing. This research was designed to evaluate the requirements and prototype the algorithms for an automated change detection system at landscape level using various geospatial data sources including multisensor remotely sensed imagery and ancillary data layers. In this paper, the requirements for an operational change information extraction system and its associated techniques are discussed. These techniques are included in three subsystems: automated computer image understanding, multisource data fusion, and database updating and visualization.

Merging SPOT XS and SAR imagery using the wavelet transform method to improve classification accuracy

SPOT XS imagery is often used to generate land use categories. Foresters, urban planners, environmental engineers, environmental scientists, and other researchers then use the resulting land use categories. State of the knowledge research in remote sensing attempts to improve the accuracy of classification techniques. This study examines the effectiveness of the wavelet technique for the fusion of SAR (synthetic aperture radar) and SPOT XS imagery before classification. The wavelet technique least distorts the special characteristics of the SPOT XS imagery while combining with the SAR imagery. A maximum likelihood algorithm is then employed for the fused data and SPOT XS data alone. Comparison using the same accuracy assessment data reveals differences in classification. Complementary characteristics of the SPOT XS and SAR imagery improve the classification accuracy.

Automated image registration for change detection from Landsat Thematic Mapper imagery

Image registration is a critical operation in digital change detection to correct for spatial image-to-image displacement. One challenging problem in this area is the automated registration with higher levels of accuracy. In this paper, the authors explore the basic elements for an automated image registration system based on image segmentation. The technique used region boundaries and strong edges as the matching primitives, and chain-code correlation as the similarity function. A mapping between the two images is formed by interpolation and used to resample the images. The method is automatic and computationally efficient.

An object-based approach to automated image matching

An object-based algorithm for automated image matching is proposed. Working on the objects (closed edges) detected from images, the authors develop a new method for determination of region correspondence using combined criteria of moment invariant distance and chain code correlation. Each object is first represented by moment invariants and improved chain codes that are affine-invariant features describing the shape of the objects. Region matching is then implemented in feature space and sequentially in image space. In feature space, minimum distance classification is used to identify the most robust control points for initial image resampling. In image space, region-to-region correspondence is established by the root-mean-square-error rule. The technique developed has significant implications in an operational context.