2004 Vol. 29, No. 4
National economic mobilization is a significant step when a country is faced with warfare or menace of warfare. It is momentous to research how to improve the efficiency of national economic mobilization. This paper probes the mode to achieve the software system on it in the era of information, and presents the corresponding solution in three parts: distributing database of resource based on Oracle, analyzing and decision-making system based on GIS, and document management based on Domino/Notes.
A method for airplane pose measurement is proposed. During airplane flying, optical-electric phototheodolite has been used to track airplane and at same time acquire image sequence. Based on these images, a method for airplane pose measurement from image sequence is deeply studied from the view of photogrammetry and projective geometry. A novel idea, Parapoint Feedback Iteration, is been presented at first time and used in airplane pose measurement. Experiment results show its high accuracy and reliability.
This paper proposes a new concept of remote sensed image, which can be considered as the mapping mode of the spectral space for ground features to the projective space that is two-dimensional imaging space. This concept can express well the texture information in multi-spectral and super-dimensional images, as well as that in single band image or black-white images. In order to analyze the texture information in multi-spectral and super-dimensional images, a necessary way is the coding for spectral vector of every pixel with all the spectral information. The texture of multi-spectral and super-dimensional images is based upon that kind of coding. Several approaches have been proposed for the coding of spectral vector: the coding based on the spectral similarity, the coding based on the density analysis of spectral space, the coding based on the image analysis of principal components, etc. The result of texture classifications based on one way of coding has been given out to show the effectiveness of the proposed method.
In order to hide secrete information in remote sensing image, this paper proposes a practical blind algorithm. Firstly, we segmente and extract the secrete information in remote sensing image, then make gray value supplement in the area in which the secrete information has been extracted, and produce the disguised remote sensing image in which the secrete information is wiped off. Then we use for reference the idea of information hiding and utilize the algorithm of odd-even embedding to embed the secrete sub-image imperceptibly into the disguised remote sensing image to produce the disguised one in which there hides secrete sub-image. In addition, during the course of extracting secrete information and resuming the remote sensing image, this algorithm does not need the original remote sensing image and is a blind one. Experimental results show that this algorithm is not only quite transparent and has a good effect on large amount of secrete information hiding, but also has a strong robustness against such familiar image processes as JPEG lossy compression, noise adding, scaling and cropping. Furthermore, this algorithm has no influence on various applications of the disguised remote sensing image with the hidden secrete information.
The objective of image fusion is to generate high-resolution multispectral images that can preserve the spectrum characteristics of the original of multispectral data. But the current fusion methods cause much spectrum distortion on multispectral images, which affects the accuracy of the further processing and analyses. In this paper, a new image fusion method based on image resolution degradation model is proposed to improve spatial and spectral resolution of synthetic images. In order to evaluate the spatial and spectral effects of the new method, we use IKONOS panchromatic and multispectral images as the test data, and some quantitative and qualitative measures are applied to assess the quality of fused images. The results show that the new method can keep the same spatial resolution almost as the panchromatic images, and the spectral effect of the new method is almost the same as that of wavelet based fusion methods.
This paper analyzes the relationship between MODIS-NDVI and AVHRR-NDVI of the same region at different time, based on Histogram and Feather Space methods. The results show that their shapes are the same, but MODIS NDVIs have more sensitivity to vegetation, and the range of their values is wider than that of NOAA-AVHRR. They have no strong correlations. It is not a good idea to use AVHRR NDVI directly as MODIS history NDVI.
This paper proposes a method for objects extraction based on high and color texture features. The existing methods using two-dimensional texture features are not effective enough for the extraction of the objects which are higher than their surroundings. An architecture is presented that observed points on forest and on non-forest are divided into two different groups in automatic aerial triangulation. The experimental results proves that the method of observed points divided into two different groups can improve the accuracy of automatic aerial triangulation.
This paper discusses the purposes and methods of Bayesian network structure learning, then proposes a new algorithm for this task. Based on a fully connected potential graph, we enter the expert knowledge and prior knowledge in order to reduce the query space of the structures. By using CI (conditional independence) tests, it can be pruned a fully connected potential graph to a best PG, which is expected to approximate the undirected version of the underlying directed graph. The experimental results of fault diagnosis in automobile are provided to illustrate the feasibility and efficiency of the new algorithm.
An image restoration algorithm based on the Mask technique is proposed to deal with the quality descend case of light decreasing around, asymmetry of intensity and hue, declination of contrast, image blur and noise etc. A Mask image is produced to simulate the image quality descend model with considering all the quality descend factors together. And a blocked Wallis transformation is applied to obtain the clear edge and best contrast to the Mask restoration image. The restoration experiments to three kinds of quality descending images show that the restoration technique based on the mask can remove the quality descending affection determined on the nature and position and realize image restoration without knowing the concrete quality descending function.
According to the characteristics of support vector machines (SVMs), a separability measure of feature space based on support vector data description is proposed. By using this measure, two hierarchical multi-category SVMs, binary-Tree SVMs and single layer clustering SVMs, are presented. Experimental results with multi-spectral remote sensing image show that the proposed separability measure is very effective. The experiments also indicate that the single layer clustering SVMs is substantially faster than 1-v-1 SVMs, 1-v-r SVMs and DAG SVMs in classification, while maintaining comparable accuracy to these algorithms.
This paper proposes to by using Harris operator extract the feature points of moving objects in sequence images, then combines the local information entropy with Walsh transformation to carry out the feature point matching in sequence image.
Rapidly reconstructing texture in 3DCM can be aided greatly if the automatically texture processing for land-based video becomes possible. Therefore, an algorithm for the automatic rectification of texture for building facades based on vanishing point geometry and the geometric constrains of known orientation of parallel object lines is presented. Relevant experiments are then implemented.
This paper is focused on how to extract linear feature from remote sensing image by watershed transform with the help of the information about the linear object. The method extracting the information and incorporating it into the watershed transform is detailed. The method consists of the following steps. ① The marker point of the linear object is detected from the input image; ② These points are utilized to modify the gradient of the input image, which aims to suppress the regional minimum that does not correspond to the linear object; ③ The modified gradient image is segmented by the watershed transform. The segmented regions that correspond to the linear object were merged to represent the object.
This paper presents an algorithm to find homonymous line from images based on Delaunay triangulation(TIN). The search range on right image of a given line on left image is decided from TIN. The relation coefficient between given line and selected matching line in search range of right image is computed. The line with the largest relation coefficient is homonymous line.The method is tested feasible.
For the nonlinear system Kalman filter in surveying field, this paper proposes a new linearization method, after the analysis of two common linearization methods. By this method, the error of linearization is similar to extended Kalman filter. Furthermore, the linear observation equation is the same to the least square method, so the comparison between Kalman filter and least square can be done during the data processing, which is helpful to the precision analysis and quality control of the positioning. Especially for GPS positioning, the linear filter equations are listed.
By analyzing the data obtained by GPS in dynamic deformation, wavelet analysis is used to filter GPS data of dynamic deformation, distill deformation character and separate different frequency of deformation.Then this paper puts forward the research on Kalman filtering model of dynamic deformation based on wavelet analysis, through combining wavelet analysis with Kalman filtering. The result of the united model is advantageous to enhance and improve GPS data accuracy and its effect is notable, which indicates the superiority to single deformation model.
This paper discusses the shortage of texture analysis based on pattern spectrum(PS) firstly. A new method, conditioned pattern spectrum, based on the integration of texture spectrum(TS) and PS is put forward. The experimental results show that the new method is reasonable and applicable.
This paper developes a general software to acquire arbitrary motion from a video sequence. A general 3D articulate human model with changeable size, color and surface shape is prepared. This model is then driven either automatically or manually to match with the moving body in the consecutive image frames. This matching starts from key frames that contain key poses. A specific orientation interpolation is used to get the motion status in in-between frames and finally a smooth motion is obtained. This method is suitable for personal use to meet wide needs of human motion acquisition.
A novel technical way to solve the problem of high resolution images positioning is proposed. The novel method utilizes only available small-scale topographic map and multi-resolution images without Ground Control Points or large scale topographic map or GPS and IMU data. The first step is to perform rectification and positioning of lowest resolution image by small-scale topographic map, wherein the matching of corresponding linear objects is explained and the algorithm is developed. Then, the image pyramid is to perform image matching layer by layer by Least Square Matching Method until the highest resolution image at the image pyramid top is rectified and positioned. Differential rectification is also discussed to reach the conclusion that DEM come from small-scale topographic map meets the requirement of planar positioning of high spatial resolution images.
By using wavelet theory, texture measure can be presented in different scales and details. The method gives us a new way of image recognition on texture. This paper introduces the wavelet analysis method for extracting the texture of SAR image. Four kinds of image texture samples are selected from aerocraft SAR image, three kinds of wavelet transform (3/5 waveform lifting, 7/9 waveform lifting, biorthorgonal wavelet) are used in the dyadic wavelet transform for SAR image, and minimum distance classification method based on wavelet character coefficients is used to classify image texture. The experiment shows that the target textual information can be effectively extracted by using wavelet transform.
On the basis of raster data structure of GIS, directed by oil and gas reservoir formation theory, and using the GIS modeling method of surface water flow direction analysis for reference, this paper designs a special GIS-based algorithm to simulate oil and gas migration pathways, and proposes a GIS-based oil and gas migration pathway modeling flow. As a case of application, the deep-buried oil and gas migration pathways has been simulated successfully in some basin of northern China.