This paper proposes a new unsupervised and fully automatic method to detect the boundaries in color natural images, inspired in the human visual model proposed by Grossberg. One of the hypotheses of Grossberg, the FAC...
详细信息
ISBN:
(纸本)9783642310744;9783642310751
This paper proposes a new unsupervised and fully automatic method to detect the boundaries in color natural images, inspired in the human visual model proposed by Grossberg. One of the hypotheses of Grossberg, the FACADE, admits complementary specialized streams at the bifurcation of the parvocellular pathway in the visual cortex: one of the branches performs edge processing and the other performs surface processing. In a similar way, this proposal has two parallel processes that are integrated at the end. The edge processing is implemented through a classical edge-detection method, whereas the surface processing is performed through a region growing method. The proposed integration scheme eliminates false contours resulted from the region growing guided by the result of edge detection, and eliminates the noise resulted from the edge detection as well, now guided by the result of the region growing, thus taking advantage of their complementary natures. Experiments on a large set of color images show that the results of the proposed system are closer to the human perception than the those correspondent to the individual methods (each branch), in quantitative and qualitative terms.
To achieve submillimetric spatial resolution, a new detection block has been designed for the LabPET ii, a small animal PET scanner being developed at Universite de Sherbrooke. Each detection block consists of 2 array...
详细信息
ISBN:
(纸本)9781467310826;9781467310833
To achieve submillimetric spatial resolution, a new detection block has been designed for the LabPET ii, a small animal PET scanner being developed at Universite de Sherbrooke. Each detection block consists of 2 arrays of 4x8 avalanche photodiodes (APD) individually coupled to an 8x8 scintillator array, to form 64 independent and parallel DAQ channels. This new detection block entails an 8-fold increase in pixel density compared to the LabPET T I. A 64-channel mixed-signal Application Specified Integrated Circuit (ASIC) was designed to extract relevant PET data in real time from the LabPET ii detection blocks. The ASIC is expected to support up to 3000 PET events/sec per channel. In order to interface the ASICs forming the PET camera with the storage units, a real-time FPGA-based digital DAQ system was designed. The DAQ system allows event harvesting, processing and transmission to a distant computer for image reconstruction as well as system programming and calibration. Real-time event processing embedded in the DAQ includes energy computation using a time-over-threshold (TOT) conversion scheme, timing corrections and event sorting trees. A real-time coincidence engine analyzes events and only keeps relevant information to minimize data throughput and post-acquisition data processing. The architecture consists of 3 layers of FPGA-based electronics wired through gigabit links: a Front-End board extracts timing and energy along with a pixel address, a Hub board sorts incoming events chronologically and a Coincidence board matches coincident events and copes with randoms estimation. Every FPGA in the different layers is accessible through an Ethernet link. The real-time digital architecture sustains the required throughput of similar to 111 Mevents/s for a similar to 37000 channels scanner configuration.
An object-based imageprocessing technique is applied to detect inundated areas using Land sat images. An interoperable Web-based system was developed to conduct the analyses so that redundant steps in Land sat image ...
详细信息
An object-based imageprocessing technique is applied to detect inundated areas using Land sat images. An interoperable Web-based system was developed to conduct the analyses so that redundant steps in Land sat imageprocessing can be effectively eliminated. A review process is used to discover and develop suitable algorithms to automatically detect inundated areas and immediately transfer the results to the Web-based interface. This is a significant improvement over currently available methods for inundation detection systems. The tool is expected to be a practical inundated area detection function and be applicable across wide-reaching areas.
In this thesis, we develop statistical methods for extracting significant information from biomedical signals. Biomedical signals are not only generated from a complex system but also affected by various random factor...
In this thesis, we develop statistical methods for extracting significant information from biomedical signals. Biomedical signals are not only generated from a complex system but also affected by various random factors during their measurement. The biomedical signals may then be studied in two aspects: observational noise that biomedical signals experience and intrinsic nature that noise-free signals possess. We study Magnetic Resonance (MR) images and speech signals as applications in the one- and two-dimensional signal representation. In MR imaging, we study how observational noise can be effectively modeled and then removed. Magnitude MR images suffer from Rician-distributed signal-dependent noise. Observing that the squared-magnitude MR image follows a scaled non-central Chi-square distribution on two degrees of freedom, we optimize the parameters involved in the proposed Rician-adapted Non-local Mean (RNLM) estimator by minimizing the Chi-square unbiased risk estimate in the minimum mean square error sense. A linear expansion of RNLM's is considered in order to achieve the global optimality of the parameters without data-dependency. parallel computations and convolution operations are considered as acceleration techniques. Experiments show the proposed method favorably compares with benchmark denoising algorithms. Parametric modelings of noise-free signals are studied for robust speech applications. The voiced speech signals are often modeled as the harmonic model with the fundamental frequency, commonly assumed to be a smooth function of time. As an important feature in various speech applications, pitch, the perceived tone, is obtained by way of estimating the fundamental frequency. In this thesis, two model-based pitch estimation schemes are introduced. In the first, an iterative Auto Regressive Moving Average technique estimates harmonically tied sinusoidal components in noisy speech signals. Dynamic programming implements the smoothness of the fundamental
Nowadays, single-chip cache-coherent multi-cores up to 100 cores are a reality and many-cores of hundreds of cores are planned in the near future. This technological shift undertaking by the high-end computer-industry...
详细信息
Nowadays, single-chip cache-coherent multi-cores up to 100 cores are a reality and many-cores of hundreds of cores are planned in the near future. This technological shift undertaking by the high-end computer-industry is converging with the design motivation of other domains like embedded and HPC industries. In this paper, we propose to investigate the scalability of the same four unmodified, shared-memory, image and signal processing oriented parallel applications on two targets: (i) embedded - TSAR, a single-chip 256-cores based, Cycle-Accurate-Bit-Accurate simulated, cc-NUMA many-core; and (ii) high-end - an AMD Opteron Interlagos, 64-core based, cc-NUMA many-core. Beside our scalability results on both cc-NUMA targets, our contributions include two operating system mechanisms: (i) a distributed, client/server based, scheduler design allowing the kernel to offer scalable inter-threads synchronization mechanisms; and (ii) a kernel-level memory affinity technique named Auto-Next-Touch allowing the kernel to transparently and automatically migrate physical pages in order to enforce the locality of thread's memory accesses. Although these two mechanisms are implemented and evaluated in ALMOS (Advanced Locality Management Operating System) running on the TSAR target, they remain applicable to other shared-memory operating systems.
In this work, we introduce a new approach for the signal deconvolution problem, which is useful for the enhancement of neutron radiography projections. We attempt to restore original signals and get rid of noise prese...
详细信息
In this work, we introduce a new approach for the signal deconvolution problem, which is useful for the enhancement of neutron radiography projections. We attempt to restore original signals and get rid of noise present during acquisition or processing, due to gamma radiations or randomly distributed neutron flux. Signal deconvolution is an ill-posed inverse problem, so regularization techniques are used to smooth solutions by imposing constraints in the objective function. Various popular algorithms have been developed to solve such problem. This paper proposes a new approach to the nonlinear degraded signals restoration which is useful in many signal enhancement applications, based on a synergy of two swarm intelligence algorithms: particle swarm optimization (PSO) and bacterial foraging optimization (BFO) applied for total variation (TV) minimization, instead of the standard Tikhonov regularization method. We attempt to reconstruct or recover signals using some a priori knowledge of the degradation phenomenon. The truncated singular value decomposition and the wavelet filtering methods are also considered in this paper. A comparison between several powerful techniques is conducted.
Membrane computing is a novel class of distributedparallel computing models. Application of membrane computing to imageprocessing is an interesting topic. In this paper, we propose a thresholding method based on P s...
详细信息
Existing cloud frameworks involve isolating low-level operations within an application for data distribution and partitioning. This limits their applicability for many applications with complex data dependency conside...
详细信息
Existing cloud frameworks involve isolating low-level operations within an application for data distribution and partitioning. This limits their applicability for many applications with complex data dependency considerations. This paper aims to explore new methods of partitioning and distributing data in the cloud by fundamentally re-thinking the way in which future data management models will need to be developed on the Internet. Loosely-coupled associative computing techniques, which have so far not been considered, can provide the break-through needed for a distributed data management scheme. Using a novel lightweight associative memory algorithm known as Edge Detecting Hierarchical Graph Neuron (Edge HGN), data retrieval/processing can be modeled as a pattern recognition problem, conducted across multiple records within a single-cycle, utilizing a parallel approach. The proposed model envisions a distributed data management scheme for large-scale data processing and database updating that is capable of providing scalable real-time recognition and processing with high accuracy while being able to maintain low computational cost in its function.
Adaptive quantization matrix design for different block sizes is one of the possible methods to improve the RD performance in video coding and has recently attracted the focus of many researchers. In this paper, we fi...
详细信息
Adaptive quantization matrix design for different block sizes is one of the possible methods to improve the RD performance in video coding and has recently attracted the focus of many researchers. In this paper, we first analyze the shortcomings of the evenly distributed distortion method which was proposed recently. In order to tackle these problems, we propose two modified methods, method I with relaxed distortion constraints and method ii is iterative boundary distortion minimization problem considering variance adaptively. Both problems can be solved using convex optimization effectively and efficiently. Simulations have been conducted based on HM4.0, which is the reference software of the latest High Efficiency Video Coding (HEVC). Simulation results show the effect of our proposed methods. Both methods show their significance when evaluated by RD performance.
The processing of microscopic tissue images and especially the detection of cell nuclei is nowadays done more and more using digital imagery and special immunodiagnostic software products. Since several methods (and a...
详细信息
The processing of microscopic tissue images and especially the detection of cell nuclei is nowadays done more and more using digital imagery and special immunodiagnostic software products. Since several methods (and applications) were developed for the same purpose, it is important to have a measuring number to determine which one is more efficient than the others. The purpose of the article is to develop a generally usable measurement number that is based on the “gold standard” tests used in the field of medicine and that can be used to perform an evaluation using any of image segmentation algorithms. Since interpreting the results themselves can be a pretty time consuming task, the article also contains a recommendation for the efficient implementation and a simple example to compare three algorithms used for cell nuclei detection.
暂无评论