 |
 | Fusion of Noisy Multi-sensor Imagery
Author : Mishra, Anima;Rakshit, Subrata
Source : Defence Science Journal ; Vol:58(1) ; 2008 ; pp 136-146
Subject : 621.38 Electronics
Keywords : Image fusion;Edge maps;Information maps;Noise filtering;Multi-resolution;Wavelets;Laplacian pyramids;Multi-sensor images
Abstract : Interest in fusing multiple sensor data for both military and civil applications has been growing. Some of the important applications integrate image information from multiple sensors to aid in navigation guidance, object detection and recognition, medical diagnosis, data compression, etc. While, human beings may visually inspect various images and integrate information, it is of interest to develop algorithms that can fuse various input imagery to produce a composite image. Fusion of images from various sensor modalities is expected to produce an output that captures all the relevant information in the input. The standard multi-resolutionbased edge fusion scheme has been reviewed in this paper. A theoretical framework is given for this edge fusion method by showing how edge fusion can be framed as information maximisation. However, the presence of noise complicates the situation. The framework developed is used to show that for noisy images, all edges no longer correspond to information. In this paper, various techniques have been presented for fusion of noisy multi-sensor images. These techniques are developed for a single resolution as well as using multi-resolution decomposition. Some of the techniques are based on modifying edge maps by filtering images, while others depend on alternate definition of information maps. Both these approaches can also be combined. Experiments show that the proposed algorithms work well for various kinds of noisy multisensor images. |
| | Algorithm for fast registration of radar images
Author : Rakshit, Subrata ;Deodhare, Dipti
Source : Defence Science Journal ; Vol:52(3) ; 2002 ; pp 243-251
Subject : 621.396.9 Radars;621.39 Telecommunication
Keywords : Radars ;Image processing ;Radar imagery ;Guidance
Abstract : "Radar imagery provides an all-weather and 24 h coverage, making it ideal for critical defence applications. In some applications, multiple images acquired of an area need to be registered for further processing. Such situations arise for battlefield surveillance based on satellite imagery. The registration has to be done between an earlier (reference) image and a new (live) image. For automated surveillance, registration is a prerequisite for change detection. Speed is essential due to large volumes of data involved and the need for quick responses. The registration transformation is quite simple, being mainly a global translation. (Scale and rotation corrections can be applied based on known camera parameters). The challenge lies in the fact that the radar images are not as feature-rich as optical images and the image content variation can be as high as 90 per cent. Even though the change on the ground may not be drastic, seasonal variations can significantly alter the radar signatures of ground, vegetation, and water bodies. This necessitates a novel approach different from the techniques developed for optical images. An algorithm has been developed that leads to fast registration of radar images, even in the presence of specular noise and significant scene content variation. The key features of this approach are adaptability to sensor/terrain types, ability to handle large content variations and false positive rejection. The present work shows that this algorithm allows for various cost-performance trade-offs, making it suitable for a wide variety of applications. The algorithm, in various cost-performance configurations, is tested on a set of ERS images. Results of such tests have been reported, indicating the performance of the algorithm for various cost-performance trade-offs." |
| | Object Hierarchy-based Supervised Characterisation of Synthetic Aperture Radar Sensor Images
Author : Rishabh, Ish;Rakshit, Subrata
Source : Defence Science Journal ; Vol:58(1) ; 2008 ; pp 159-167
Subject : 621.38 Electronics
Keywords : Object segmentation;Object characterisation;Image characterisation;Terrain-based matching;SAR imagery
Abstract : A method of supervised characterisation of synthetic aperture radar (SAR) satellite images has been discussed in which simple object shape features of satellite images have been used to classify and describe the terrain types. This scheme is based on a multilevel approach in which objects of interest are first segmented out from the image and subsequently characterised based on their shape features. Once all objects have been characterised, the entire image can be characterised. Emphasis has been laid on the hierarchical information extraction from the image which enables greater flexibility in characterising the image and is not restricted to mere classification. The paper also describes a method for giving relative importance among features, i.e., to give more weights to those features that are better than others in distinguishing between competing classes. A method of comparing two SAR sensor images based on terrain elements present in the images has also been described here. |
|
