Diabetic retinopathy is one of the leading causes of preventable blindness. Screening programs using color fundus photographs enable early diagnosis of diabetic retinopathy, which enables timely treatment of the disea...
详细信息
ISBN:
(纸本)9781479946037
Diabetic retinopathy is one of the leading causes of preventable blindness. Screening programs using color fundus photographs enable early diagnosis of diabetic retinopathy, which enables timely treatment of the disease. Exudate detection algorithms are important for development of automatic screening systems and in this paper we present a method for detection of exudate regions in color fundus photographs. The method combines different preprocessing and candidate extraction algorithms to increase the exudate detection accuracy. First, we form an ensemble of different candidate extraction algorithms, which are used to increase the accuracy. After extracting the potential exudate regions we apply machine learning based classification for detection of exudate regions. For experimental validation we use the DRiDB color fundus image set where the presented method achieves higher accuracy in comparison to other state-of-the art methods.
Logo search is widely required in many real-world applications. As a special case of near-duplicate images, logo pictures have some particular properties, for instance, suffering from flipping operations, e.g., geomet...
详细信息
ISBN:
(纸本)9781479957521
Logo search is widely required in many real-world applications. As a special case of near-duplicate images, logo pictures have some particular properties, for instance, suffering from flipping operations, e.g., geometry-inverted and brightness-inverted operations. Such operations completely change the spatial structure of local descriptors, such as SIFT, so that image search algorithms based on Bag-of-visual-Words (BovW) often fail to retrieve the flipped logos. We propose a novel descriptor named Max-SIFT, which finds the maximal SIFT value sequence for detecting flipping operations. Compared with previous algorithms, our algorithm is extremely easy to implement yet very efficient to carry out. We evaluate the improved descriptor on a large-scale Web logo search dataset, and demonstrate that our method enjoys good performance and low computational costs.
The Mojette Transform is a mathematical transform that has approximately 20 years of research history. The transform is the exact definition of the Discrete Radon transform. Despite of the simplicity of its mathematic...
详细信息
The Mojette Transform is a mathematical transform that has approximately 20 years of research history. The transform is the exact definition of the Discrete Radon transform. Despite of the simplicity of its mathematics, this transform is a really computationally intensive procedure. The purpose of this paper is to introduce algorithms based on the Mojette Transform that the research group (composed from three university members) developed in recent years. These algorithms are mainly made for purposes of imageprocessing, which are increasingly popular in fields where the real-time operation is of utmost importance. This is why was developed an algorithm that can reach nearly the runtime of real-time imageprocessing needs. For the implementation the LabvIEW (trademark of National Instruments) was used. The benefits of using LabvIEW are explained, and the details of the new algorithm, that is different from the previous implementations presented in [12] and [13], are given. Finally partial results of the implementation of the proposed algorithms are discussed.
In this paper, a course of low complexity Fast Mode Decision (FMD) algorithms as well as the corresponding hardware architecture is presented. Firstly, the depth information of co-located block from previous frame is ...
详细信息
ISBN:
(纸本)9781479957521
In this paper, a course of low complexity Fast Mode Decision (FMD) algorithms as well as the corresponding hardware architecture is presented. Firstly, the depth information of co-located block from previous frame is used to predict the size of current block. Then, for a certain sized block, the inter prediction residual is analyzed to determine whether to terminate current check or to skip some unnecessary modes and split to smaller size. Finally, the corresponding hardware architecture is proposed based on state machine mechanism. Simulation results show that these proposed algorithms reduce the encoding time by 40.8～70.3%, without incurring any noticeable performance degradation. Hardware synthesis results demonstrate that the proposed architecture achieves a max frequency of about 193 MHz.
An improved Hough Transform based fingerprint alignment approach is presented, which improves computing time and memory usage with accurate alignment parameter (rotation and translation) results. This is achieved by s...
详细信息
An improved Hough Transform based fingerprint alignment approach is presented, which improves computing time and memory usage with accurate alignment parameter (rotation and translation) results. This is achieved by studying the strengths and weaknesses of existing Hough Transform based fingerprint alignment algorithms, and combining the strengths to an improved approach. The results of alignment parameters are checked manually for each image after alignment. The experimental results indicated that the improved approach improves accuracy at a less computing time and memory usage. The experiments were performed using FvC2000 and FvC2004 databases as there are different impressions of fingerprints that represent possible fingerprint rotations and translations.
Undergraduate engineering students who are learning Digital Signal processing (DSP) are expected to have the ability to implement their theoretical knowledge in various applications soon after graduation. In this pape...
详细信息
ISBN:
(纸本)9781479946037
Undergraduate engineering students who are learning Digital Signal processing (DSP) are expected to have the ability to implement their theoretical knowledge in various applications soon after graduation. In this paper, we present a laboratory experiment developed for undergraduate students that addresses the challenge of getting them familiar with implementing DSP algorithms in heterogeneous multicore systems. In a top-down approach, the students first gain control of the development environment, and then implement DSP algorithms on a general purpose and on a digital signal processor core. Through the experiment, they get to appreciate the advantages of DSP core architecture in performing signal processingalgorithms, and learn methods for timing and data transfer between cores while meeting real-time constraints. In a limited time frame, this hands-on laboratory experiment exposes the students to state-of-the-art multicore development practices and increases their knowledge and interest in DSP and in embedded programming.
Improving the users' Quality of Experience (QoE) in modern 3D Multimedia systems is a challenging proposition, mainly due to our limited knowledge of 3D image Quality Assessment algorithms. While subjective QoE me...
详细信息
ISBN:
(纸本)9781479957521
Improving the users' Quality of Experience (QoE) in modern 3D Multimedia systems is a challenging proposition, mainly due to our limited knowledge of 3D image Quality Assessment algorithms. While subjective QoE methods would better reflect the nature of human perception, these are not suitable in real-time automation cases. In this paper we tackle this issue from a new angle, using deep learning to make predictions on the user's QoE rather than trying to measure it through deterministic algorithms. We benchmark our method, dubbed Quality of Experience for 3D images through Factored Third Order Restricted Boltzmann Machine (Q3D-RBM), with subjective QoE methods, to determine its accuracy for different types of 3D images. The outcome is a Reduced Reference QoE assessment process for automatic image assessment and has significant potential to be extended to work on 3D video assessment.
Although the perceptual image hashing is one of the promising techniques for image authentication, most existing methods cannot well distinguish content changing manipulations from acceptable, content preserving modif...
详细信息
ISBN:
(纸本)9781479957521
Although the perceptual image hashing is one of the promising techniques for image authentication, most existing methods cannot well distinguish content changing manipulations from acceptable, content preserving modifications, especially when the size of the manipulated area is relatively small. In this regard, a new image hash algorithm is proposed to enhance the tamper detection capability by employing one of the most well-known local feature descriptors, Histogram of Oriented Gradients (HOG), for the feature extraction method. In this paper, image intensity transform using a random number generator, HOG feature computation, Successive Mean Quantization Transform (SMQT), and bit-level permutation are utilized to obtain a secure and robust hash value. Additionally, the performance of the proposed method is measured, and compared with existing algorithms by the Receiver Operating Characteristics (ROC) analysis.
Primal-dual interior-point methods are used in imageprocessing to solve inversion problems that can be reduced to constrained convex minimization. Such iterative methods require the solution of a sequence of linear s...
详细信息
Primal-dual interior-point methods are used in imageprocessing to solve inversion problems that can be reduced to constrained convex minimization. Such iterative methods require the solution of a sequence of linear systems which are used to derive descent directions. This approach is very consuming in terms of computing time and memory usage for large-scale problems, unless the involved matrices have a specific structure. This is the case in the spectral unmixing problem where these matrices are block-diagonal when no spatial regularization is considered. Here, we consider the edge-preserving regularized case and we propose to tackle the linear system solving using a majorization-minimization (MM) approach based on separable quadratic majorant functions. The resulting systems have the same structure as in the non-regularized case and can thereby be solved efficiently. The interior-point algorithm is speeded-up while remaining convergent. An example of spectral unmixing is proposed to illustrate the efficiency of this approach.
暂无评论