Showing papers on "Change detection published in 1991"

Change detection in SAR imagery

Abstract: The two major difficulties associated with SAR image change detection are addressed. These are the removal of speckle noise and the registration of information between images. The problem of image registration is severe in airborne SAR imagery due to the unpredictable nature of the aircraft track. Although inertial navigation systems may be employed to measure this motion, the accuracy obtained is insufficient to allow the creation of large rectilinear images. However, autofocus techniques are used here to measure residual aircraft motions thus allowing the production of large geometrically accurate images. The second problem of speckle reduction has been approached in two ways. The first technique applies an intensity segmentation algorithm to each image and the regions generated by this segmentation are then compared using the change detection algorithm. An alternative approach to the problem of speckle removal is to use neural network methods to learn the speckle removal and region generation ...

Change Detection in Sar Images

Abstract: A system has been developed which is able to detect changes in an area by using two satellite SAR images taken from similar tracks, but from different passes. Accurate registration is performed by correlation on a subpixel basis. The change detection then involves taking the ratios rather than differences. Different ratio masks are used to detect different features. Threshold values used to produce the final change detection image, is dependent on the ratio mask size. Thresholds are estimated by calculating the probabilty of false alarm (pfa) for different threshold values until pfa I 1 % for the ratio image. Tests have been carried out using two SEASAT SAR images.

Change Detection From Remotely Sensed Multi-temporal Images Using Spatial Segmentation

Abstract: A new method is proposed for detection of temporal change between remotely sensed multi-spectral images. This method uses regression and segmentation techniques for reduction of apparent change due to noise and/or variation of recording gain and offset between two images. In the method, the images are normalized by fitting a line to the whole pixel scattergram between them. The line is specified by slope and intercept. Temporal change is detected when the hypothesis is rejected that a line fitted to the whole pixel scattergram is identical to that to a local area scattergram. The hypothesis is statistically tested. Noise is reduced by above regression process. The spatial segmentation is used for division of the image into small areas where density scattergrams are described well by a single set of slope and intercept. In this paper, we describe the principle and the procedures of this method. The validity of this method is confirmed by numerical simulation. A result obtained by application of this method to a pair of actual multi-temporal multi-spectral images is also shown.

A unified framework for statistical change detection

Abstract: The authors address the problem of detecting abrupt changes in the statistical properties of signals and dynamical systems, which is of key interest for failure detection and monitoring. They present a unified parametric statistical framework for both the design and the performance analysis of change detection algorithms. This framework is introduced in the simplest case of change in the scalar parameter of the distribution of an independent random sequence. Situations of increasing complexity are investigated, namely additive changes in linear regression and state-space models and spectral changes in both scalar and multidimensional cases. The link with the analytical redundancy and detection filter approaches is also stressed. A general tool for performance evaluation is introduced: the average run length function, which can be used for estimating both the mean time between false alarms and the mean delay for detection, and for tuning the algorithms. >

Television signal recorder and television signal reproducing device

Abstract: PURPOSE:To record and reproduce lots of television signals efficiently. CONSTITUTION:The device is provided with an A/D converter section 11, a motion detection section 12 detecting a motion signal of a video image formed based on an inter-frame difference circuit, a moving area size measurement section 13 measuring the size of a moving area in a frame from the obtained motion detection signal, an inter-frame difference arithmetic operation section 14 obtaining an inter-frame difference of the obtained moving area size signal, a comparator section 16 obtaining a scene change detection signal when the obtained inter-frame difference exceeds a reference level set in advance to a level setting section 15, a still picture extract section 2 extracting a still picture at that point of time from the video signal every time the scene change detection signal is received, and a recording section 4 recording the extracted still picture data, its time code and an audio data sequentially as each unit.

Double talk detecting system in echo cancellor

Abstract: PURPOSE:To speedily detect a double talk state without being influenced by means of echo route fluctuation by directly identifying double talk and the fluctuation of an echo route. CONSTITUTION:When the echo route is fluctuated and a change on the time average of the values of YjXj(m) occurs, one of mean value calculation circuits 22 and comparators 23 in a correlative mean shift detection circuit 20a detect the change. A deciding circuit 30a receives the detection output of a power change detection circuit 10a and the detection output of the correlative mean shift detection circuit 20a and judges it to be the fluctuation of the echo route when the correlative mean shift detection circuit 20a detects the change of a mean value, and judges it to be double talk when the power change detection circuit 10a detects the change of power and the correlative mean shift detection circuit 20a does not detect the change of the mean value. Thus, a processing that double talk is detected by identifying it from the fluctuation of the echo route is executed.

Multi-sensor fusion for classification and change-detection in remote-sensed imagery

Abstract: A single remote-sensed image provides a snap shot of a scene at a particular instance in time. The measurement process, and hence the grey-levels of the final image, is dependent on the properties of the scene, the sensor and the sensing conditions. Combining information from multiple images, possibly at different times or with different sensors, enables further properties of the scene to be deduced. The motivation for this is to either improve classification accuracy of the scene, or else to quantify temporal variations in either shape or attributes of the scene. This latter problem is termed change detection. The approach described is based on feature-based correspondence of regions in the two input images. Images are processed from pixel data to be stored in a vector database, via a segmentation algorithm. Single-image classification and attribute extraction may be carried out prior to combination.

Method and device for detecting change of image

Abstract: PURPOSE:To detect the remarkable change of an image content with high accuracy by processing an input digital image signal by using correlation in an image once, finding the statistical distribution of a processed result, and checking the temporal change of the statistical distribution. CONSTITUTION:A processing means 30 which processes the input digital image signal by using the correlation between picture elements in each of continuous images, and a statistical distribution calculation means 32 which calculates the statistical distribution of the processed result from a processed signal are provided. Also, a collation means 37 which outputs a change detection signal to collate the statistical distribution of the processed result in the present image with that of the processed result in a past image and to represent the change of the statistical distribution by referring to the former and the latter is provided. In such a way, it is possible to detect the remarlable change between two continuous images with high accuracy.

Detection of Abrupt Changes in Statistical Models

Abstract: : This dissertation investigates different types of disorder problems by using sequential procedures for on-line implementation The problem is considered within the framework of detecting abrupt changes in an observed random process when the disorder can occur at unknown times The focus of this work is on quickest detection methods for cumsum procedures implemented for different parametric and nonparametric nonlinearities and their performance evaluation Both the non-Bayesian (Maximum-Likelihood) and the Bayesian frameworks are presented but the focus is mainly on non-Bayesian methods for which detailed analysis is provided The use of Brownian motion approximations is also included and provides an additional viewpoint of analyzing the performance for both the non-Bayesian and Bayesian methods

Real-time SAR change-detection on a parallel computer architecture

Abstract: Synthetic aperture radar (SAR) provides a coherent, high-resolution, imaging system. The author outlines the principles underlying the implementation of a real-time SAR processor capable of generating linear focussed imagery by combining motion compensation, together with autofocus and phase correction. Finally, the changes between the SAR image and a previous version are detected to indicate significant objects in the scene. These algorithms have been previously developed and demonstrated on a serial general-purpose computer. In considering their transfer onto a parallel architecture based on the i860 Meiko Computing Surface the authors identified the practical restrictions imposed by processor speed, local memory and communications bandwidth, and demonstrated that, generally, the local memory is likely to be the dominant limitation. He proposes a possible architecture which balances the limiting factors as far as possible and meets the requirement for generating fully focussed, undistorted, SAR images with change detection at a real-time rate. >

A Hybrid Methodology for Detecting Cartographically Significant Features Using Landsat TM Imagery

Abstract: : A general Change Detection (CD) methodology is investigated that involves a hybrid mix of image processing, spectral transformation, and statistical pattern recognition techniques. The Hybrid Methodology attempts to combine various forms of supporting and conflicting evidence for change into a resulting change map. The approach involves differencing registered multiband scene pairs that have undergone a spectral transformation, generating threshold masks, and applying a classifier to the masked multiband scene pairs.

An Automated Software System for Updating Digital Terrain Databases from All-Source Imagery. Phase 1

Abstract: : VEXCEL Corporation's Phase l SBIR research effort for U.S. Army's Engineer Topographic Laboratories concentrated on the feasibility of creating the primary tools for the prototype development in Phase II of a digital change detection work station. This system in intended to be capable of detecting long-term (6 months to 1 year) and/or seasonal changes from all-source imagery. The system is intended to be hosted on a SUN-4 platform operating under a UNIX/C software environment. The emphasis of the present Phase l effort was on the two major technical challenges for the development of such a system: precision image registration and robust change detection and analysis. Most of this effort was directed toward automated SAR-optical image registration and automated change cueing experiments. Change cueing is an initial step in change detection for identifying regions where possible change events may have occurred. The automated registration and cueing efforts were successful over the data sets tested. These data sets did not contain appreciable terrain-induced distortions. Theoretical improvements for algorithms are recommended for addressing such terrain-induced registration complications in Phase II. Recommendations are also made for the inclusion of interactive capabilities to supplement the automated registration techniques.

State change detection system for input data

Abstract: PURPOSE:To reduce the load on a CPU and to speed up detection as to whether or not there is a state change by sampling data of plural input ports and detecting whether or not there is the state change through a circuit without the intervention of a CPU. CONSTITUTION:Buffers 41 and 4N receive data D1 and DN from input ports I1 and IN individually and a counter 2 specifies an address signal ADR1 to a common RAM 1; and a control circuit 3 generate a chip select CS and write/ read control signals EB1 and EB2 for the buffers 41 and 4N and common RAM 1, and sample data in data units of the input ports are written in the common RAM 1. A comparing circuit 5 inputs current data DATA1 of sample data of the data D1 and DN from the buffers 41 and 4N and last data DATA2 from the common RAM 1 to detect whether or not there is the state change of data, and sends out an interruption processing request IRQ to a CPU 6 when detecting the state change. Consequently, the load on the CPU is reduced and the detection of the state change of the data of the input ports is speeded up.

Change Detection In SPOT Satellite Images

Abstract: The rapid growth of the remote sensing and observation data has created the need for automatic image analysis systems. The photo-interpreter remains a key element of such systems. Repetitive, often trivial tasks should be left to the automatic system; the operator detains the expertise and the responsibility for Ehange interpretation.

Supervisory and controlling equipment

Abstract: PURPOSE:To respond to the change of traffic quantity by providing a state change detection processing part, and switching the detecting method of state change. CONSTITUTION:When monitor information is inputted from a low-order device 15, the state change detection processing parts 11-13 perform the detecting processing of a state with a system set respectively. A control information conversion part 28 generates control information 31, 32 by converting control information 29 supplied from a high-order device 27, and information representing the traffic quantity of communication are included in the control information 31, 32, and a state change detection processing selection part 18 selects the state change detection processing part 11 when low traffic quantity exists, and the state change detection processing part 12 when high traffic quantity exists, and the state change detection processing part 13 when medium traffic quantity exists. Thereby, it is possible to make the congestion of a line hard to generate even when high traffic quantity exists.

An Attempt of Automatic Change Detection for Map Revision

Abstract: Necessity of an automatic change detection method was emphasized for efficient map revision. The significance and problems of utilizing existing old maps for a change detection system were discussed. A couple of problems in applying basic image understanding techniques to change detection were reviewed, and an attempt was made to solve them with a prototype system developed for use with satellite images and map data.The attempt clarified that every automatic change detection system outputs some errors in its results, and that such systems must have some efficient error inspection functions incorporated well with human inspection capability in order to be practically used. An example of such inspection systems was also presented.