Synthetic Aperture Radar (SAR) images are efficient and reliable source of information in extraction of damaged regions in case of floods. In assessment of damage accurately due to floods, image registration of optica...
详细信息
ISBN:
(纸本)1595930361
Synthetic Aperture Radar (SAR) images are efficient and reliable source of information in extraction of damaged regions in case of floods. In assessment of damage accurately due to floods, image registration of optical (before-flood) and SAR images (after-flood) has to be carried out efficiently. In this paper, we discuss a robust multi-sensor image registration algorithm using scale invariant feature points for keypoint extraction. For matching the keypoints, a multi-objective genetic algorithm is developed with angle, distance and vicinity criterion as the fitness functions. This optimization process helps in matching the scale invariant feature points. From the obtained results, the performance of the image registration is evaluated and it is concluded that the proposed approach is efficient. Copyright 2014 ACM.
We present a GPU-based implementation of the saliency model proposed by Achanta et al. [1] to perform real-time and detailed saliency map generation. We map all the components of the algorithm to GPU-based kernels and...
详细信息
ISBN:
(纸本)1595930361
We present a GPU-based implementation of the saliency model proposed by Achanta et al. [1] to perform real-time and detailed saliency map generation. We map all the components of the algorithm to GPU-based kernels and data structures. The parallel version of the algorithm is able to accurately simulate the desired results in a very low time. We describe the streaming pipeline and address many issues in terms of obtaining high throughput on multi-core GPUs. We highlight the parallel performance of the algorithm on three different generations of GPUs. On a high-end NVIDIA Tesla K20m, we observe up to 600x order of magnitude performance improvement as compared to a singlethreaded CPU-based algorithm, and about 300x order of magnitude improvement over a CPU-based OpenCV implementation. Copyright 2014 ACM.
Simulation and rendering of material weathering is of great interest in computergraphics as it facilitates modeling of realistic objects in virtual scenes. We present a physiochemically based methodology to model pit...
详细信息
ISBN:
(纸本)1595930361
Simulation and rendering of material weathering is of great interest in computergraphics as it facilitates modeling of realistic objects in virtual scenes. We present a physiochemically based methodology to model pitting corrosion. Our simulation has two components. The first models pit nucleation on the surface of the object. It is implemented as a stochastic process, exponentially deteriorating with time. The second component models pit growth using another stochastic process inuenced by pit characteristics (geometry, neighborhood and exposure). We validate our model with measurements published in the corrosion literature and then demonstrate its effectiveness in generating realistic images. Our corrosion model leads to shape degradation. Color degradation is less tractable. Hence we instead learn that directly from acquired images of corroded models. We present an efficient rendering algorithm that generates color, normal and depth textures to display the corrosion of any object of interest. This is an important step in linking the appearance to structural damage. Copyright 2014. ACM.
Range-image super-resolution has evolved in recent years to improve the images acquired by low-resolution range-cameras. In this regard, some local filtering based approaches are quite popular as they achieve reasonab...
详细信息
ISBN:
(纸本)1595930361
Range-image super-resolution has evolved in recent years to improve the images acquired by low-resolution range-cameras. In this regard, some local filtering based approaches are quite popular as they achieve reasonable quality range-images while maintaining high computational efficiency. In this work, we propose a novel and improved local approach, which is inspired by the popular Guided image Filtering method, that employs information from an associated color image for the task of range-image super-resolution. Our approach accounts for consideration of the content of both color image and range image explicitly, to drive the enhancement process. We show that our filter reduces noise for noisy range-images along with better edge enhancement, especially for higher up-sampling factors. Our experimentation also demonstrate that our approach performs better than other prominent local filtering approaches both in terms of depth precision and spatial resolution without any considerable increase in computational time. Copyright 2014 ACM.
In this paper, we propose a region based saliency detection algorithm using a total variation based regularizer. We aim to obtain salient objects that are uniformly highlighted. The use of the regularizer facilitates ...
详细信息
ISBN:
(纸本)1595930361
In this paper, we propose a region based saliency detection algorithm using a total variation based regularizer. We aim to obtain salient objects that are uniformly highlighted. The use of the regularizer facilitates the removal of textures from the image. This leads to an image that contains piecewise constant gray-valued segments. This texture-free image is sparsely segmented into a small number of regions using the expectation maximization algorithm assuming a Gaussian mixture model. We compute saliency of regions using their intensity and spatial features. The saliency map is then thresholded to obtain the salient regions of the image. Next we employ an image matting technique to extract the exact boundaries of the salient objects from the image. This approach leads to noise-free saliency maps containing uniformly highlighted salient regions. The experimental comparison with existing saliency detection algorithms demon-strates the superiority of the proposed technique. Copyright 2014 ACM.
This paper proposes a framework for classification of label-free, unstained, leukemia cell lines MOLT and K562 in microuidics based Imaging Flow Cytometry (IFC). These two cell lines differ in their internal cell comp...
详细信息
ISBN:
(纸本)1595930361
This paper proposes a framework for classification of label-free, unstained, leukemia cell lines MOLT and K562 in microuidics based Imaging Flow Cytometry (IFC). These two cell lines differ in their internal cell complexity in an IFC image. Each cell is localized by finding a closed cell membrane binding internal organelles. An existing non-iterative graph based contour detection algorithm is extended and is effectively used to segment out the cells. Features reecting the size, circularity and internal cell complexity are extracted and used for classification using linear Support Vector Machine. Copyright is held by the authors.
In this paper, a method for color image segmentation using multiscale intuitionistic fuzzy roughness measure is proposed. The traditional roughness measure tends to over focus on the little important homogeneous regio...
详细信息
ISBN:
(纸本)1595930361
In this paper, a method for color image segmentation using multiscale intuitionistic fuzzy roughness measure is proposed. The traditional roughness measure tends to over focus on the little important homogeneous regions but is not accurate enough to measure the homogeneity in an image. By applying the theories of scale space and using intuitionistic fuzzy representation for images, roughness is measured under multiple scales. Multiscale representation can tolerate the disturbance of trivial regions, and intuitionistic fuzzy representation deals with hesitancy in image boundary, therefore produces precise segmentation results. Copyright 2014 ACM.
This paper presents an efficient system for palm-dorsa vein pattern based recognition system. It can handle efficiently the problem of false palm-dorsa veins which can be created by many ways such as ink, tattoos, art...
详细信息
ISBN:
(纸本)1595930361
This paper presents an efficient system for palm-dorsa vein pattern based recognition system. It can handle efficiently the problem of false palm-dorsa veins which can be created by many ways such as ink, tattoos, artificial vein pattern paper fixed on the palm-dorsa. Hand-dorsa images acquired under visible and infrared lights are used. Since vein pattern from infrared light has spurious and genuine vein pattern, spurious vein pattern is removed from it by using vein pattern from visible light. It has been tested on 600 visible and 600 infrared hand-dorsa images. Experimental results indicate that the proposed system performs efficiently. Copyright 2014 ACM.
Object tracking is a critical task in surveillance and activity analysis. One of the major issues in visual target tracking is variation in illumination. In this paper, we propose an improved weighted histogram approa...
详细信息
ISBN:
(纸本)1595930361
Object tracking is a critical task in surveillance and activity analysis. One of the major issues in visual target tracking is variation in illumination. In this paper, we propose an improved weighted histogram approach for mean shift tracking to achieve illumination invariance. The proposed scheme can effectively reduce changes in the illumination of targets with respect to background. Depending on changes in the target appearance, the target model is updated. The experimental results show that the proposed scheme can lead to faster convergence, more accurate localization and robust tracking in several illumination scenarios, when compared to existing methods. Copyright 2014 ACM.
We have developed a ground vehicle capable of maneuvering in an open environment negotiating outdoor obstacle course autonomously carrying a payload by finding colored lanes, obstructions and navigating through GPS wa...
详细信息
ISBN:
(纸本)9780819499424
We have developed a ground vehicle capable of maneuvering in an open environment negotiating outdoor obstacle course autonomously carrying a payload by finding colored lanes, obstructions and navigating through GPS waypoints. In this paper, we will be discussing the hardware components like mechanical and electrical design, various sensors used and software components of the vehicle like imageprocessing, environment mapping and navigation algorithms. The vehicle uses its sensors to develop some limited understanding of the environment, which is then used by control algorithms to determine the next action to take in the context of a human provided mission goal.
暂无评论