Chong Fan

Robust Affine Invariant Feature Extraction for Image Matching

A new approach is presented to extract more robust affine invariant features for image matching. The novelty of our approach is a hierarchical filtering strategy for affine invariant feature detection, which is based on information entropy and spatial dispersion quality constraints. The concept of spatial dispersion quality is introduced to quantify the spatial distribution of features. Moreover, an integrated algorithm combined by the filtering strategy, maximally stable extremal region (MSER) and scale invariant feature transform, is introduced for affine invariant feature extraction. Since Mikolajczyk et al. identified that MSER is the best detector in many cases, we design an experiment to compare our approach (ED-MSER) with the standard MSER. By using two stereo pairs and an image sequence with different types of imagery, the experiment indicates that ED-MSER can always get much higher repeatability and matching score compared to the standard MSER and other algorithms, thus benefiting the subsequent image matching and many other applications.

POCS Super-Resolution Sequence Image Reconstruction Based on Improvement Approach of Keren Registration Method

This paper introduces the keren sub-pixel registration method and point out its disadvantage. Moreover, this paper put forward a new improvement approach about keren method and its iterative method. The improvement approach takes the four parameters affine transformation model instead of the rigid body transformation model. This change avoids the error that is brought on by the tailor series expansion of angle and improves the precise of image registration greatly. The experiment shows that the improvement approach makes less absolute error of angle than keren method and its iterative algorithm. The improvement approach makes the absolute error of translation parameters under 0.1 pixels in the case of the rotation angel of 15 degree and under 0.01 pixels in the case of the small rotation angle using our pictures. At last, the projection onto convex set (POCS) method is used to reconstruct high-resolution image from several low-resolution image sequences. As a result, we find that the reconstruction algorithm based on our improvement registration approach has better effect than the reconstruction algorithm based on keren iterative registration method

An improvement approach based on keren sub-pixel registration method

This paper introduces the keren sub-pixel registration method and point out its disadvantage. Moreover, this paper put forward a new improvement approach about keren method and its iterative method. The improvement approach base on the four parameters affine transformation model and abandon the rigid body transformation model. This change avoids the error that is brought on by the tailor series expansion of angle and improves the precise of image registration greatly. The experiment show that the improvement approach makes less absolute error of angle than keren method and its iterative algorithm in the case of great noise pollution. The improvement approach can make the absolute error of translation parameters under 0.1 pixels in the case of the rotation angel of 15 degree and under 0.01 pixels in the case of the small rotation angle

Multi-frame Image super-resolution based on knife-edges

In this paper, a new method of super-resolution is proposed. Instead of the traditional regularization term of the blur function such as TV regularization and Tichonov, new regularization of the blur function is constructed by using the knife-edge in the low resolution images. The proposed method can reduce the overestimation of the traditional methods. Other two previous super-resolution methods have been taken into the comparison to the proposed method, and results of both synthetic data and real data have shown a much better performance(both in PSNR(peak signal-to-noise ratio) and visional effects) of the proposed method than the other methods.

Slant edge method for point spread function estimation

The traditional edge method (TEM) is applicable only in areas whose knife edges are horizontal or vertical. The slant edge method (SEM) requires using the approximate function in the edge spread function (ESF) fitting. The ESF sample is uneven and therefore affects the accuracy of the estimated point spread function (PSF). This study proposes a novel slant edge method (NSEM), including a knife-edge area extraction algorithm that ensures the evenness of the ESF sample and a PSF estimation algorithm that directly fits the line spread function, to improve the accuracy of the estimated PSF. The validity and applicability of the NSEM have been proven in theory and practice, respectively. Compared with the TEM and SEM, the NSEM exhibits a clear improvement in accuracy and stability across the three experiments conducted in this study.

Building Change Detection Using Old Aerial Images and New LiDAR Data

Building change detection is important for urban area monitoring, disaster assessment and updating geo-database. 3D information derived from image dense matching or airborne light detection and ranging (LiDAR) is very effective for building change detection. However, combining 3D data from different sources is challenging, and so far few studies have focused on building change detection using both images and LiDAR data. This study proposes an automatic method to detect building changes in urban areas using aerial images and LiDAR data. First, dense image matching is carried out to obtain dense point clouds and then co-registered LiDAR point clouds using the iterative closest point (ICP) algorithm. The registered point clouds are further resampled to a raster DSM (Digital Surface Models). In a second step, height difference and grey-scale similarity are calculated as change indicators and the graph cuts method is employed to determine changes considering the contexture information. Finally, the detected results are refined by removing the non-building changes, in which a novel method based on variance of normal direction of LiDAR points is proposed to remove vegetated areas for positive building changes (newly building or taller) and nEGI (normalized Excessive Green Index) is used for negative building changes (demolish building or lower). To evaluate the proposed method, a test area covering approximately 2.1 km2 and consisting of many different types of buildings is used for the experiment. Results indicate 93% completeness with correctness of 90.2% for positive changes, while 94% completeness with correctness of 94.1% for negative changes, which demonstrate the promising performance of the proposed method.

Improved fast mean shift algorithm for remote sensing image segmentation

Image segmentation plays a crucial role in object-based remote sensing information extraction. This study improves the existing mean shift (MS) algorithm for segmenting high resolution remote sensing imagery by adopting two strategies. First, a pixel-based, fixed bandwidth and weighted MS algorithm is applied to cluster the image. In this process, the space bandwidth is selected according to the resolution of remote sensing images, and the range bandwidths of each band are calculated based on grey feature and the plug-in rule. Gaussian kernels are used for clustering. Second, a region-based MS algorithm is applied to globally merge modes which are obtained in the first step. The spatial and range bandwidths are adaptively adjusted based on the clustering result of the first step. Experimental results with two Quickbird images show that the improved algorithm is superior to the typical MS algorithm, producing high precision and requiring less operation time.

Projections onto Convex Sets Super-Resolution Reconstruction Based on Point Spread Function Estimation of Low-Resolution Remote Sensing Images

To solve the problem on inaccuracy when estimating the point spread function (PSF) of the ideal original image in traditional projection onto convex set (POCS) super-resolution (SR) reconstruction, this paper presents an improved POCS SR algorithm based on PSF estimation of low-resolution (LR) remote sensing images. The proposed algorithm can improve the spatial resolution of the image and benefit agricultural crop visual interpolation. The PSF of the high-resolution (HR) image is unknown in reality. Therefore, analysis of the relationship between the PSF of the HR image and the PSF of the LR image is important to estimate the PSF of the HR image by using multiple LR images. In this study, the linear relationship between the PSFs of the HR and LR images can be proven. In addition, the novel slant knife-edge method is employed, which can improve the accuracy of the PSF estimation of LR images. Finally, the proposed method is applied to reconstruct airborne digital sensor 40 (ADS40) three-line array images and the overlapped areas of two adjacent GF-2 images by embedding the estimated PSF of the HR image to the original POCS SR algorithm. Experimental results show that the proposed method yields higher quality of reconstructed images than that produced by the blind SR method and the bicubic interpolation method.

POCS super-resolution sequence image reconstruction based on image registration excluded aliased frequency domain

This paper introduces the theory of super-resolution image reconstruction and degraded model in brief, and presents a new super-resolution image reconstruction algorithm. The algorithm bases on the new image registration excluded aliased frequency domain and the Projection Onto Convex Set (POCS) method. The algorithm can precisely estimate the image registration parameter by excluding aliased frequency domain of the low-resolution images and killing the center part of the magnitude spectrum. In order to compute the shifts and the rotation angle, we set up the polar coordinates in the center of the image. By computing the frequency function of the rotation angle by integrating over radial lines, the algorithm converts the two-dimension correlation to one-dimension correlation. And then, the POCS method is used to reconstruct high-resolution image from these aliased image sequences. As a result, we find that the reconstruction algorithm has the same precision of image registration as the spatial image registration and good effect of super-resolution image reconstruction.

A Three-Dimensional Transmission Line Management System Based on TerraExplorer Pro

Aiming to the features and demands of transmission line information management and merging the remote sensing images, digital elevation model and GIS layers, this paper creates a three-dimensional scene of transmission line, which takes TerraExplorer Pro as basic platform and combines the attribute information of transmission line, and develops an interactive and visualized management system which is oriented to power system. The system realizes the 3D display, query and analysis of transmission line and provides the integrated supports to the design, construction, operation and maintenance at a certain extent