Showing papers on "Change detection published in 1998"

Speaker, Environment and Channel Change Detection and Clustering via the Bayesian Information Criterion

Abstract: In this paper, we are interested in detecting changes in speaker identity, environmental condition and channel condition; we call this the problem of acoustic change detection. The input audio stream can be modeled as a Gaussian process in the cepstral space. We present a maximum likelihood approach to detect turns of a Gaussian process; the decision of a turn is based on the Bayesian Information Criterion (BIC), a model selection criterion well-known in the statistics literature. The BIC criterion can also be applied as a termination criterion in hierarchical methods for clustering of audio segments: two nodes can be merged only if the merging increases the BIC value. Our experiments on the Hub4 1996 and 1997 evaluation data show that our segmentation algorithm can successfully detect acoustic changes; our clustering algorithm can produce clusters with high purity, leading to improvements in accuracy through unsupervised adaptation as much as the ideal clustering by the true speaker identities.

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

Abstract: 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%.

Thresholding for change detection

Abstract: Image differencing is used for many applications involving change detection. Although it is usually followed by a thresholding operation to isolate regions of change there are few methods available in the literature specific to (and appropriate for) change detection. We describe four different methods for selecting thresholds that work on very different principles. Either the noise or the signal is modelled, and the model covers either the spatial or intensity distribution characteristics. The methods are: 1) a Normal model is used for the noise intensity distribution, 2) signal intensities are tested by making local intensity distribution comparisons' in the two image frames (i.e. the difference map is not used), 3) the spatial properties of the noise are modelled by a Poisson distribution, and 4) the spatial properties of the signal are modelled as a stable number of regions (or stable Euler number).

An unsupervised change detection and recognition system for forestry

Abstract: This article presents a new unsupervised method (AutoChange) for change detection and identification. It uses, as an input, two images, acquired on different dates, and a parameter list given by the user. Change detection and identification are performed in separate procedures, and the output is a five channel image estimating the change magnitude and characterizing the changed and unchanged areas. The method carries out the change analysis using homogeneous units selected from the images and only in the ultimate phase the whole image is classified. Changes are detected and identified using clustering in two phases. First, clustering is performed on the earlier and later images to form the so called 'primary clusters'. Second, clustering is performed within the primary clusters of the later image to produce the 'secondary clusters'. Then the change magnitude and change type are obtained by comparing the primary clusters in the earlier image to the secondary clusters in the later image. The method...

Temporal video segmentation using unsupervised clustering and semantic object tracking

Abstract: This paper proposes a content-based temporal video segmentation system that integrates syntactic (domain- independent) and semantic (domain-dependent) features for auto- matic management of video data. Temporal video segmentation in- cludes scene change detection and shot classification. The proposed scene change detection method consists of two steps: detection and tracking of semantic objects of interest specified by the user, and an unsupervised method for detection of cuts, and edit effects. Object detection and tracking is achieved using a region matching scheme, where the region of interest is defined by the boundary of the object. A new unsupervised scene change detec- tion method based on two-class clustering is introduced to eliminate the data dependency of threshold selection. The proposed shot classification approach relies on semantic image features and ex- ploits domain-dependent visual properties such as shape, color, and spatial configuration of tracked semantic objects. The system has been applied to segmentation and classification of TV programs col- lected from different channels. Although the paper focuses on news programs, the method can easily be applied to other TV programs with distinct semantic structure. © 1998 SPIE and IS&T. (S1017-9909(98)00803-4)

Speckle filtering in satellite SAR change detection imagery

Abstract: Repeat-pass Synthetic Aperture Radar (SAR) imagery is useful for change detection. A disadvantage of SAR is the system-inherent speckle noise. This can be reduced by filtering. Various filter types and methods are described in the literature, but not one fits the speckle noise in change detection imagery. A new method is therefore developed in this paper. The new method is based on filtering the logarithmic-scaled ratio of SAR images. Logarithmic scaling changes the multiplicative speckle noise in the ratio-image into additive noise and alters the distribution, which simplifies and optimizes the subsequent filter process. The filter in the new method consists of an additive LLMMSE filter (Kuan et al. 1985), preceded by a structure detection stage for a better contour preserving performance. Testing the new method on a repeat-pass satellite SAR image-set gave an accurate overview of changes compared to a colour-composite of both images, other optical remote sensing images and maps of the same area.

Statistical change detection with moments under time-varying illumination

Abstract: In this paper, an illumination-independent statistical change detection method is proposed. The proposed method consists of two parts. First, based on our defined circular shift moments, structural changes can be distinguished from those due to time-varying illumination in the noise-free case. Moreover, the amount of computation is less than that of the shading model method. Second, in the light of the characteristics of the defined moments, a statistical decision rule is also proposed to cope with the effects of noise. The change detection problem can be treated as one of hypothesis testing. Critical values can be chosen according to the desired level of significance. Experimental results indicate that the proposed method detects changes accurately in the time-varying illumination case.

Change detection study of Kuwait City and environs using multi-temporal Landsat Thematic Mapper data

Abstract: Landsat Thematic Mapper (TM) images acquired on 17 February 1986, and 28 February, 1993, respectively, were used to demonstrate their usefulness for both surface/spectral mapping and temporal change detection in Kuwait City and environs. High pass spatial filters with a relatively large kernel (201 pixels by 201 pixels) were used to enhance high frequency information in both the bright desert and dark urban areas. The filtered results were edge enhanced to sharpen the local textural information. Colour composites were made for analyses using the TM bands-2, -3, -4 and-2, -4, -7. The two multi-temporal images were geometrically and radiometrically calibrated to each other and used as input to an automatic change detection procedure. The 'change image' composite, made from the individual change image results generated using TM bands -2, -4, and-7, detected and mapped temporal changes dealing with urban development, vegetation, coastal wetlands, and sand sheet surface differences caused by the large oil spil...

An analysis of changes in a multi-date data set: A case study in the Ikitelli area, Istanbul, Turkey

Abstract: The analysis of changes has become an important use of multi-spectral Landsat TM data. With the repetitive acquisition of imagery, it is possible to determine the types and extent of changes in an environment. Many digital change-detection techniques, such as image overlay, image differencing, and principal component analysis, have been used widely for this purpose. The objective of this study was to detect land cover changes by using the multi-date Landsat TM imagery for the Ikitelli area, Istanbul, Turkey, employing different methods. Each change detection method used was assessed, with its ability to detect specific changes.

Content-based video abstraction

Abstract: The paper addresses automatic scene change detection, key frame selection, and similarity ranking which constitute the main steps of a content based video abstraction system. Unlike other methods, the proposed algorithm performs scene change detection and key frame selection in one step. We treat scene change detection as a two-class classification problem and employ automatic threshold selection techniques originally developed for image binarization. A quantitative measure for retrieval of similar scenes according to their color content is also defined. The described scheme can be applied to both uncompressed and MPEG compressed video, and can be implemented in real time. Performance of the algorithm has been analyzed on real TV sequences, and comparison with some previously introduced techniques are provided.

A Study on Automatic Shot Change Detection

Abstract: We present a study of various automatic shot change detection methods for video segmentation which have been proposed in the literature. We identify representatives of the main approaches to the problem of shot cut detection and compare them experimentally on a large number of sequences. We discuss the relative merits of various algorithms and their uses.

Robust video shot change detection

Abstract: We present a novel improvement to existing schemes for abrupt shot change detection. Existing schemes declare a shot change whenever the frame to frame histogram difference (FFD) value is above a particular threshold. In such an approach, a high value for the threshold results in a small number of false alarms and a large number of missed detections while a low value for the threshold decreases the number of missed detections at the expense of increasing the false alarms. We attribute this situation to the fact that the FFD cannot be reliably used as the sole indicator for the presence of a shot change. In the proposed method a two-step shot detection strategy is used which selectively uses a likelihood ratio (computed directly from the frames and not from the histograms) to confirm the presence of a shot change. Such a two-step checking increases the probability of detection without increasing the probability of false alarm. The improvement proposed is simple and computationally cheap. Tests with a wide variety of video sequences prove the efficacy of the proposed approach.

Video scene change detection using the generalized sequence trace

Abstract: We propose an algorithm to detect scene changes in a video sequence in the compressed domain. We define a feature vector extracted from each frame that we call the generalized trace. We examine various ways of processing the generalized trace to determine the temporal location of scene changes in a video stream.

Object detection and tracking in time-varying and badly illuminated outdoor environments

Abstract: A real-time method for object detection and tracking in outdoor environments where illumination can be very low and not constant is presented. A hierarchical (two levels) change detection method is employed to detect moving objects in the scene. At the first level, a focusof- attention stage is applied to individuate image areas containing moving objects; at the second level, each selected image area is inspected at higher accuracy to improve the detection probability and to obtain an accurate binary reconstruction of the object shape. A background updating procedure is used to adapt the background image to the changes in the scene. Then, an extended Kalman filter is applied to track multiple objects entering the scene. Results are reported on real scenarios in the presence of shadows, occlusions, light reflections, and bad environmental conditions.

A fast scene change detection on MPEG coding parameter domain

Abstract: This paper proposes a very fast and accurate scene change detection algorithm on MPEG coding parameter domain. The fast operation can be obtained by spatio-temporally subsampling coding information and by exploiting only coding parameters extracted in variable length decoding (VLD) stage, while the accurate detection is accomplished by examining the statistical characteristics of various scene changes on the coding parameter domain. The computer simulation shows that the proposed algorithm can accomplish detection more than 5 times faster than that of real-time playback for MPEG-2 video sequences using a standard workstation. It is also shown that most of abrupt scene changes, dissolve transitions, and wipe transitions have been successfully detected.

Hierarchical scene change detection in an MPEG-2 compressed video sequence

Abstract: In this paper, we propose an efficient scene change detection algorithm for direct processing of MPEG-2 video bitstreams. The proposed algorithm utilizes the hierarchical structure of the compressed bitstreams and statistical characteristics of the coded parameters, thus greatly reducing computational requirement compared to pixel domain processing with full decompression. Occurrence of scene change is checked first in a GOP level, and if the result is affirmative it is checked again in lower levels: sub-GOP and each picture. We used several metrics for different levels: variance of DC images for I-pictures, number of macroblock types for P-pictures and motion vector types for B-pictures.

Matching and aligning features in overlayed coverages

Abstract: 1. ABSTRACT The problems caused by locational error when overlaying spatial data from different sources have been recognised for some time, and much research has been directed towards finding solutions. In this paper we present a solution in the form of an algorithm that seeks to match and align semantically equivalent features prior to overlay. It is assumed that, because of locational error, semantically equivalent features will not always be geometrically equivalent. The technique has been developed to assist in the detection of change between multi-date vector-defined data sets. Initial results, obtained by applying our algorithm to land cover data, are presented. 1.1

Automated procedures for integration of satellite images and map data for change detection: the archangel project.

Abstract: The need for automation in the registration of image to image and image to map is widely recognised and work has been going on for some time in both the photogrammetric and remote sensing disciplines. For large scale aerial photography the automation of absolute orientation is tied in with feature recognition but for satellite data operations can take place at a lower level and the problem of relief displacement and occlusions is less. The image to image registration problem is some way to being solved but the image to map problem is much more difficult. The main bottleneck is the identification of points which can be seen on an image and on the corresponding map. This is a manual task at present requiring a considerable amount of time. It is particularly difficult in areas where there is little cultural development and where poor, out of date maps are the only ones available. The ARCHANGEL project is designed to develop a system for the automatic registration of satellite data to maps and this paper describes the system developed and the some of the results achieved. A second difficult and tedious task is the identification of areas where change has taken place. The job of the skilled interpreter is to determine the nature of change, s/he does not wish to spend time finding where the change has occurred. The key aspect of the project is therefore to develop a method which will automatically find common points on the maps and images in any type of area and which will also find areas where change has occurred. The paper will describe the algorithms which have been developed and tested and integrated into a prototype system in an environment called XPECT, developed by Earth Observation Sciences Ltd. Results to date show that the basic algorithms give good results for segmentation, map processing and matching. The algorithms have been tested by the developers and the overall system will be tested by Swedish Space Corporation as end users.

On fuzzy clustering and content based access to networked video databases

Abstract: Video databases and video on demand represent an important application of the evolving global information infrastructure. However, video querying involves a lot of user interaction and feedback based query refinement, which can generate large traffic volumes on the network if full video segments are sent. To aid in efficient video browsing, search and retrieval across the network, we need to find good compact representations for long video sequences. Representative frames (Rframes) provide such a representation. Extant algorithms use scene change detection to segment video into shots and pick Rframes. However, scene change detection techniques fail badly in presence of gradual scene changes which are quite prevalent in most videos. We present another way of finding Rframes using fuzzy clustering without dealing with any scene change detection algorithms. Fuzzy clusters provide a more natural approach to this problem since membership of a frame in some particular scene is not binary. This allows us to handle gradual scene changes. We report on our approach, present preliminary experimental results, and discuss ongoing work.

Robust scene-change detection in MPEG compressed domain

Abstract: Video is an important and challenging medium and requires sophisticated indexing schemes for efficient retrieval from visual databases. Video segmentation is a fundamental step in video indexing and involves detection of scene changes. In this paper, we propose a fast and robust algorithm for detecting video shot boundaries in the MPEG-2 compressed bit stream with minimal decoding.

Land cover change detection with change vector in the red and near-infrared reflectance space

Abstract: An enhanced land cover change indicator product is produced using the two 250-m spatial resolution bands of the moderate resolution spectroradiometer (MODIS) of the NASA Earth Observing System. The rationale for creating the 250-m resolution land cover change product is that a very high proportion of land cover changes occur at the finest MODIS spatial resolutions. Multiple change detection algorithms are employed for the product generated because different algorithms detect different types of change. Among these algorithms there are two using the change vector in the red and near-infrared reflectance space. An early example of using change vector analysis for change detection is in Malila (1980). A more recent example is Johnson and Kasischke (1998). In these examples the general concepts of multispectral change vector analysis were described and applied to detect vegetation changes for the cases in specific locations and seasons. In this paper, the change vector in the red and near-infrared (NIR) reflectance space is analyzed for different types of relevant land cover changes. The algorithms using the change vector characteristics obtained with NOAA's Advanced Very High Resolution Radiometer (AVHRR) data to detect the changes are then described. Validations of these two algorithms with the simulated MODIS data from Landsat Thematic Mapper (TM) images of two different areas are presented.

Potential of an airborne laser scanner system for change detection of urban features and orthoimage development

Abstract: The Kobe earthquake in 1995 demonstrated an urgent need of developing a change detection mapping system of buildings with high temporal accuracy to save people from damaged buildings. An airborne laser scanning system (ALSS) is considered to have the potential of meeting such a need of detecting building changes in 3D shape because of its capability of direct measurement of ground elevation and short turnaround time. This study employed the data acquired by ALSS at three different occasions to detect changes of buildings in 3D shape. Images generated from a simple processing of these ALSS data clearly showed building changes without omission errors, the detection of which, otherwise, would require intensive manual airphoto interpretation. The combination of a CCD array sensor with the laser scanning system also indicated the potential of ALSS for automated orthoimage development, even though the current system developed in this study still needs further improvement in image displacement correction.

Scene change detection in MPEG domain

Abstract: Video is an important and challenging media and requires sophisticated indexing schemes for efficient retrieval from visual databases. Video segmentation is a fundamental step in video indexing and involves detection of scene changes. In this paper, we propose a fast and robust algorithm for detecting video shot boundaries in the MPEG-2 compressed bitstream with minimal decoding.

A fast effective scene change detection and adaptive rate control algorithm

Abstract: The rate control algorithm of Test Model 5 (TM5) can not handle scene change properly, so the visual quality is consequently worsened. A fast effective scene changes detection method without much additional computation is proposed and an adaptive rate control algorithm is given, which not only guarantees the buffer not to overflow or underflow, but also compensates the visual degradation at scene change point and keeps consistent visual quality. Simulation results show that this algorithm can effectively improve the visual quality when scene change occurs.