The study of substances with a crystal structure is a complex multi-step process. The key step in the crystalline substance analysis is the unit cell parameter estimation. The estimation of the crystal lattice unit ce...
详细信息
Currently, robots are increasingly being used in every industry. One of the most high-tech areas is creation of completely autonomous robotic devices including vehicles. The results of various global research prove th...
详细信息
Currently, robots are increasingly being used in every industry. One of the most high-tech areas is creation of completely autonomous robotic devices including vehicles. The results of various global research prove the efficiency of vision systems in autonomous robotic devices. However, the use of these systems is limited because of the computational and energy resources available in the robot device. The paper describes the results of applying the original approach for imageprocessing on reconfigurable computing environments by the example of morphological operations over grayscale images. This approach is prospective for realizing complex imageprocessingalgorithms and real-time image analysis in autonomous robotic devices.
Single-pixel camera (SPC) is a low-cost compressive imaging architecture that obtains random projections of the scenes using binary coded apertures. After the acquisitions, image reconstructions are usually obtained b...
详细信息
Single-pixel camera (SPC) is a low-cost compressive imaging architecture that obtains random projections of the scenes using binary coded apertures. After the acquisitions, image reconstructions are usually obtained by nonlinear and relatively expensive optimization-based algorithms. Recent works have focused on designing the binary coded apertures to improve the speed of the reconstruction algorithms and the sampling complexity of the compressed sensing systems. However, it has been shown that image recovery is not necessary for image classification from compressive measurements, where only specific features of the images are required. This work proposes a deep learning approach for image classification directly from SPC measurements. In this approach, a neural network is trained to simultaneously learn the linear binary sensing matrix and the non-linear classification parameters, considering the constraints imposed by the SPC. Specifically, the first layer learns the sensing matrix, and subsequent layers perform the classification directly on the compressed measurements. Simulation results from two image datasets validate the proposed method, which provides the best classification accuracy along with a binary sensing matrix.
In this paper, we propose a novel fuzzy logic data association algorithm to resolve the problem of visual multi-object tracking. First of all, a fuzzy logic data association approach is proposed to calculate the assoc...
In this paper, we propose a novel fuzzy logic data association algorithm to resolve the problem of visual multi-object tracking. First of all, a fuzzy logic data association approach is proposed to calculate the association probabilities between the objects and observations (or detection responses) by use of a set of fuzzy if-then rules. Based on the similarities of object appearance, shape and motion models, several fuzzy inference systems are designed to confirm the fuzzy membership degrees between the objects and observations, the association probability is generated by weighted affinities, where the weight of the affinity is the output of fuzzy logic. Finally, the experiment results on several public sequences demonstrate that the proposed algorithm has superiorities over other comparatively leading tracking algorithms in terms of accuracy and robustness.
Cancer is a lethal disease if not detected in an early stage. This paper presents an outline of different types of cancers and recent advancement in soft computing techniques for their detection. It focuses on how dif...
详细信息
ISBN:
(纸本)9789811068720;9789811068713
Cancer is a lethal disease if not detected in an early stage. This paper presents an outline of different types of cancers and recent advancement in soft computing techniques for their detection. It focuses on how different imageprocessing techniques are optimized using neural networks, fuzzy logic, and genetic algorithms to detect cancers.
Modified Gram Schmidt (MGS) is one of the well-known forms of QR decomposition (QRD) algorithms. It has been used in many signal and imageprocessing applications to solve least square problem, linear equations or to ...
详细信息
Modified Gram Schmidt (MGS) is one of the well-known forms of QR decomposition (QRD) algorithms. It has been used in many signal and imageprocessing applications to solve least square problem, linear equations or to invert matrices. Nevertheless, QRD is considered a computationally expensive operation, and its sequential implementation doesn't meet the requirements of many real time applications. In this paper, we propose an optimized MGS algorithm version based on software pipelining and loop unrolling techniques. The suggested MGS version is parallel and well suited for vLIW architectures. The implementation is done under TI C6678 vLIW DSP and the obtained results show great improvements against the standard MGS and the optimized vendor QRD implementations.
image classification is one of the most important tasks in image analysis and computer vision. BP neural network is a successful classifier for the task. However, with regard to the low study efficiency and the slow c...
image classification is one of the most important tasks in image analysis and computer vision. BP neural network is a successful classifier for the task. However, with regard to the low study efficiency and the slow convergence speed in BP algorithm, some optimization algorithms have been proposed for achieving better results. Among all these methods, BP neural network improved by particle swarm optimization (PSO) and genetic algorithm (GA) may be the most successful and classical ones. Nevertheless, both GA and PSO are easy to fall into the local optimal solution, which has a great impact on the precision of classification. As a result, a novel optimization algorithm called sine cosine algorithm (SCA) is presented to improve the classification performance. The experimental results manifest that the proposed method has good performances, and the classification accuracy is better than BP neural network optimized by GA, PSO or other algorithms.
During transmission and reception the images are degraded by noise. Also images captured using different low quality devices adversely affect the visual quality of images. The presence of the noise results in loss of ...
详细信息
ISBN:
(纸本)9781538677094
During transmission and reception the images are degraded by noise. Also images captured using different low quality devices adversely affect the visual quality of images. The presence of the noise results in loss of visibility, gives a mottled, grainy, textured, or snowy appearance. Thus making the imagevisually unpleasing which drastically affects the human vision. In addition during imageprocessing, presence of noise makes it hard for further processing (impulse, Gaussian white, salt and pepper, adversarial etc.,). Existing methods use conventional filters and Neural network models for image denoising where they compromise with the visibility of image after rigorous iterations of denoising algorithms. In this paper we implement CGANs for image denoising and evaluate the performance of CGAN with different Neural network models viz., CNN,GAN for single or multiple image denoising problem. The qualitative performance of de-noised image/images is measured using PSNR and confusion matrix.
Dynamic tracking of tumor with radiation beam in radiation therapy requires prediction of real-time target location ahead of beam delivery as the treatment beam with beam or gating tracking brings in time latency. We ...
详细信息
Dynamic tracking of tumor with radiation beam in radiation therapy requires prediction of real-time target location ahead of beam delivery as the treatment beam with beam or gating tracking brings in time latency. We proposed an adaptive boosting (Adaboost) method based on the multi-layer neural network (MLP-NN) to predict the respiratory signal accuracy in our previous study. Recently, several variants of Adaboost methods showed great potential for the regression prediction problem. Hence, we investigated the prediction performance of four popular adaptive boosting method based on the MLP-NN for the respiratory prediction problem in this study. The root-mean-square-error (RMSE), correlation coefficient (CC) and maximum error (ME) between predicted and real respiratory signals, obtained from the Real-time Position Management, were used to evaluated the prediction performance. The Adaboost. RT method and the Adaboost method used in our previous study get the best RMSE and CC while the Adaboost. BCC obtained the best ME. The experiment results demonstrated that appropriate Adaboost method based on MLP-NN could predict the respiratory signal accuracy.
暂无评论