Table of Content

30 January 2015, Volume 33 Issue 1

Previous Issue    Next Issue

Signal and Information Processing

Coverage Area in Different Types of Remote Sensors:Simulation and Calculation

WU Zhao-cong1, YANG Fan1, WANG Nan1, WU Yuan1, ZHU Zhuo-wa2

2015, 33(1):  1-8.  doi:10.3969/j.issn.0255-8297.2015.01.001

Asbtract ( 1368 )   PDF (5219KB) ( 1319 )  

References | Related Articles | Metrics

The paper analyzes geometric coverage of common remote sensors for earth observation. These are classified into two modes: spherical circular coverage and spherical quadrilateral coverage. Projective geometries of the two coverage modes are induced and simulation models of coverage in nadir and side imaging modalities built. Based on the simulation models, a universal calculation process is established for geometric coverage of common remote sensors, including transformations among six ground and spatial coordinate systems. Simulations are carried out with real satellite and sensor parameters,showing effectiveness of the proposed calculation process.

Geometric Feature Representation and Building Extraction Based on Geometric Features

ZHU Jun-jie1,2, FAN Xiang-tao1,2, DU Xiao-ping1,2

2015, 33(1):  9-20.  doi:10.3969/j.issn.0255-8297.2015.01.002

Asbtract ( 1178 )   PDF (16252KB) ( 401 )  

References | Related Articles | Metrics

Problems exist in building extraction from high-resolution remote sensing images in terms of spectral features. For example, the same object has different spectra and different objects have the same spectrum, and geometric features of a building may be more typical than its spectral features. Having summarized some typical geometric features of buildings, building extraction is done based on their geometric features with features of area, fitting rectangle, length-to-width ratio, and main directions. The experiment shows that extracting buildings with geometric features is more effective than with spectral features.However, since image segmentation depends on spectral features so that calculation of geometric features is affected by spectral features, building extraction based on geometric features is not satisfactory. As the height feature of a building is the most typical geometryfeatures, introduction of height features into image segmentation and feature extraction is an effective way to improve accuracy of building extraction.
 

Segmentation of SAR Image of Rivers Based on Shearlet Transform and Krawtchouk Moment Invariants

WU Shi-hua1, WU Yi-quan1,2,3,4,5, ZHOU Jian-jiang1,MENG Tian-liang1, DAI Yi-mian1

2015, 33(1):  21-31.  doi:10.3969/j.issn.0255-8297.2015.01.003

Asbtract ( 1828 )   PDF (2316KB) ( 705 )  

References | Related Articles | Metrics

Segmentation of synthetic aperture radar (SAR) images is a key procedure in river detection and recognition. To further improve accuracy of SAR image segmentation for rivers, a segmentation method is proposed based on Shearlet transform, Krawtchouk moment invariants, and fuzzy local information C-means (FLICM). The SAR image is first decomposed with Shearlet transform, and its texture features are extracted as the first part of the feature vector. Krawtchouk moment invariants of the image are then calculated to obtain corresponding shape features used as the second part of the feature vector. Finally,
the image is clustered based on FLICM algorithm using the extracted feature vector. Thus segmentation of the river SAR image is obtained. A large number of experiments are performed. The results are compared with several recently proposed methods based on pulse coupled neural network (PCNN) combined with maximum variance ratio, and Gabor wavelet transform combined with fuzzy C-means (FCM) and FLICM clustering. It has been shown that the proposed method has clear advantages both in subjective visual effects and in terms of objective evaluation indicator such as segmentation error rate. The method can
provide river segmentation with better accuracy.
 

An Improved Strong Tracking Filtering Algorithm

QIAN Hua-ming1, GE Lei1,2, HUANG Wei1

2015, 33(1):  32-40.  doi:10.3969/j.issn.0255-8297.2015.01.004

Asbtract ( 1371 )   PDF (3119KB) ( 945 )  

References | Related Articles | Metrics

Strong tracking filtering (STF) sets small threshold to judge filtering divergence leading to fading factor with high probability, which causes excessive regulation of the filtering gain and makes the state estimation curve lack smoothness. By analyzing the operation mechanism of STF, improved STF (ISTF) is proposed. The proposed algorithm reduces probability of misjudging filter divergence by appropriately increasing the threshold. It determines the softening factor to suit different dimensions of the measurement equation, and thus avoids the disadvantages of the previous methods that determine the softening factor according to experiences. Simulation indicates that ISTF can maintain filtering accuracy and estimation smoothness, and is robust against sudden changes in the system state, showing its feasibility and effectiveness.

Numerical Study on the Relationship between Amplitudes of Ground Penetrating Radar Wave and Water Content in Soil

LI Hui-jun1,2,3, ZHONG Ruo-fei1,2,3

2015, 33(1):  41-49.  doi:10.3969/j.issn.0255-8297.2015.01.005

Asbtract ( 1513 )  

References | Related Articles | Metrics

Water content in soil can be derived from the relation between changes in ground penetrating radar signals and the soil’s dielectric constant. Using the finite difference time domain (FDTD) method, a model of layered soil is established to simulate ground penetrating radar electromagnetic waves in the soil. The transmission characteristics of layered soil are analyzed using the model. The results indicate that a functional relation exists between the soil dielectric constant and the amplitude, i.e., the wave amplitude decreases with the increased electric constant. With an indoor soil model, the obtained relationship
is verified: the computed water content agrees with the wet sensor data.
 

Reversible Data Hiding in Encrypted Images Transmitted on Demand

HAN Xi-yu, QIAN Zhen-xing, ZHANG Xin-peng, JIANG Fei

2015, 33(1):  50-58.  doi:10.3969/j.issn.0255-8297.2015.01.006

Asbtract ( 1813 )   PDF (5246KB) ( 1354 )  

References | Related Articles | Metrics

This paper proposes a new method of reversible data hiding in encrypted images. The sender down-samples the original image and calculates the residuals to generate two sets of objects: sub-images and residuals. These are encrypted with an encryption keyand sent to the server. The server applies arithmetic coding on the encrypted residuals to generate spatial redundancy, and then hides extra data using the embedding key. With the key, the receiver obtains different versions of the data transmitted by the server. Unlike the conventional methods, a transmission-on-demand mechanism is used, which can effectively reduce the amount of transmitted data on the server. With the image perfectly recovered, the data embedding rate is greatly improved.

Analysis of NDVI Changes and Its Climate Factor Drivers in Ebinur Lake Basin from 1998 to 2012

WANG Er-li, ZHOU Jun-qi

2015, 33(1):  59-69.  doi:10.3969/j.issn.0255-8297.2015.01.007

Asbtract ( 1469 )   PDF (15823KB) ( 428 )  

References | Related Articles | Metrics

To explore the impact of climate change on vegetation cover in Ebinur Lake Basin, we apply the data set SPOT VEGETATION to study the trend of normalized difference vegetation index (NDVI) and spatial pattern in this area, using a maximum synthesis and linear regression method. Combining the annual precipitation and annual average temperature in the same period, the vegetation cover driving force of climate change is discussed through partial correlation and multiple correlation analyses. The results indicate that NDVI increases significantly, showing a good development trend. Among various natural factors, precipitation has a deeper and more widespread effect on vegetation than temperature during 15 years in Ebinur Lake Basin. In addition, about 88.9% of the studyarea is impacted by non-climate factor drivers, while 11.1% is driven by climate factors and mainly distributes in the eastern basin.
 

Rate Control of I-Frames Based on URQ Model

WANG Peng, NI Cui, ZHANG Guang-yuan, LI Ke-feng

2015, 33(1):  70-78.  doi:10.3969/j.issn.0255-8297.2015.01.008

Asbtract ( 1541 )   PDF (14601KB) ( 473 )  

References | Related Articles | Metrics

In high efficiency video coding, the coding efficiency of I-frames is lower than P-frames and B-frames, causing flicker artifacts, especially in low bitrates applications. We propose a rate control algorithm for I-frames based on URQ model to improve the reconstruction quality of I-frames to suppress flicker artifacts. An I-frame is classified into three regions according to their motion vectors and complexity, termed motion region, texture-rich region and smooth region. According to different sensitivities of human visual system to different regions, visually sensitive regions are allocated more target bits. The URQ model is used to calculate QPs. Experimental results demonstrate that the proposed algorithm can efficiently suppress flicker artifacts and improve subjective and objective quality of the video.
 

Accuracy Analysis of Region Adjustment of Remote Sensing Images Based on Dual Quaternion

XIAO Hui1,2, FEI Li-jia3

2015, 33(1):  79-86.  doi:10.3969/j.issn.0255-8297.2015.01.009

Asbtract ( 1281 )   PDF (4465KB) ( 1044 )  

References | Related Articles | Metrics

According to the geometric algebra theory, a bundle adjustment method based on dual quaternion is presented. The method uses dual quaternion to describe rotation and translation of regional photo coordinates, and can linearize a strict collinearity equation. Constraint parameter adjustment is iteratively computed. The structure of normal equation in region adjustment based on dual quaternion is similar to that of traditional methods, with the required memory slightly increased with the increasing margins. The results show that, by arranging control points in four angle points and central point, the dual quaternion-based region adjustment can achieve relatively high measuring accuracy.

Forgery Detection via Inter-channel Correlation of CFA Images

ZHANG Xiao-lin, FANG Zhen, ZHANG Xin-peng

2015, 33(1):  87-94.  doi:10.3969/j.issn.0255-8297.2015.01.010

Asbtract ( 1324 )   PDF (6270KB) ( 815 )  

References | Related Articles | Metrics

Most digital cameras use a single sensor to capture only one component among the three colors (RGB) for each pixel, together with a color filter array (CFA) interpolating the other two after data acquisition. In this work, spectral correlation introduced by inter-channel CFA interpolation is exploited to realize image authentication. By analyzing differences in the frequency spectrum between inter-channel interpolated images and the natural ones, we extract forensics features from the high frequency areas of the green-red spectral difference. After re-interpolating the test image, we detect tampered images according to changes in the forensic features. Experimental results demonstrate effectiveness in forgery detection and robustness to JPEG compression.

Color Clustering Based on Chromaticity Histogram

YU Jun1, LIU Fei-hong1, WANG Zhan-feng2, YANG Jun-na1

2015, 33(1):  95-104.  doi:10.3969/j.issn.0255-8297.2015.01.011

Asbtract ( 1763 )   PDF (9315KB) ( 1310 )  

References | Related Articles | Metrics

In camouflage design, extracting the dominant color from target background is an important step. A drawback of common color clustering methods is the requirement of supervision. This paper proposes an unsupervised color clustering algorithm based on the chromaticity histogram. The chromaticity histogram is established according to the CIE 1931 system. The cluster center is automatically generated based on the distribution of pixels in the coordinate system. Euclidean distances in the chroma space are calculated one by one between pixels and each cluster center. These pixels are clustered around the
nearest clustering center. Experimental results show that the proposed clustering algorithm can accurately extract the dominant color, and automatically segment the color image. In addition, the algorithm uses less time in image segmentation than common algorithms.