جستجوی مقالات مرتبط با کلیدواژه « Image processing » در نشریات گروه « ریاضی »

It has been thirty years since Perona and Malik (PM) introduced the nonlinear diffusion equation in image processing and analysis. The problem's complexity was to find a suitable and adaptive diffusion function that smooths away noise or textures while preserving sharp edges of a sufficiently smooth intensity. This paper provides a new two-dimensional analysis of the PM diffusion equation to examine its behavior during scales and an explicit formula to select the right diffusion function adequately. In this context, we study the PM equation at the zero crossings of the first and second directional derivatives of a sufficiently smooth function in the gradient direction.

Orientation is a vital ability for humans and animals. Noticing the way insects orient in nature can be used to improve the orientation skills of robots. The main question of this research can be stated as follows. What kind of information do insects perceive of natural scenes, using their visual ability, that enables them to orient and to find the direction of movement? For orientation, the minimum of rotational image difference function (MrIDF) method can be applied using panoramic image processing [1]. In MrIDF method, even with full shift, if the distance between the location of the current view image and the reference image increases, the return path cannot be correctly identified due to the increase in the difference between the two images. In this paper, we present a solution that can be used to identify the path and return angle in places far from the reference location. We also improve the efficiency the rotIDF minimum method by using the covariance matrix adaptation evolutionary strategy (CMA-ES) optimization algorithm. We show the efficiency of this method via a navigation example. The results show that finding the direction of movement through the proposed algorithm is done with sufficient accuracy and in much less time.

Sorting agricultural products refers to grading food and other crops based on size, color, appearance, and other factors such as separating impurities, fruits, and damaged and rotten products. Today, sorting technology and related equipment for grading agricultural crops are progressing in developed countries, which can be found in most large agricultural units. Therefore, initial packaging and transportation of the product are facilitated, and more added value can be provided for farmers. This study aimed to optimize the raisin sorting machine based on a genetic algorithm to increase the quality of raisin grading. Therefore, a seedless white variety of grape samples were randomly selected and prepared from an orchard in Makan, East Azerbaijan, Iran. Digital image processing techniques such as the image processing toolbox in MATLAB were used to extract features from an image for sorting. Other meta-heuristic algorithms such as PSO, differential evolution, and artificial bee colony algorithm were used to evaluate the accuracy of the results. According to the results, the artificial bee colony algorithm had better accuracy than other algorithms, but the convergence speed was lower, and the computational volume was higher. However, the genetic and PSO algorithms had an accuracy almost equal to the artificial bee colony algorithm despite having a higher speed of convergence and lower computational operations, which can be used as the best algorithm in this application. Differential evolutionary algorithms and harmony search require processing in many iterations, and the computation time is not economical. Therefore, the clustering of raisins in industrial units requires high clustering speed and minimum error to avoid discarding or outliers, and genetics and PSO algorithms were acceptable.

This paper aims to present a useful method for magnifying images, for which it is necessary to group the pixels and define the borders. In the proposed method, images are first partitioned using suitable segmentation algorithms and then artificial neural networks (ANNs) are applied to magnify each segment individually. In the ANNs applied, training is performed using, as input, a down-sampled form of the same image to be magnified. This type of training results in a high quality zoom in each segment since the pixels in an individual segment have very close features. Evaluation results on several images verifies the higher efficiency of the proposed method than other recently developed image zooming methods.

It is no secret to anyone the essential steps of mathematics in image processing and especially in image watermarking techniques. Transformations are one of the most important mathematical tools used in image processing. It is possible to reduce the size of the image very efficiently and extract important information from it. On the other hand,  transformations enable us to move smoothly from the world of image processing (spacial domain) to the world of mathematics and matrices (frequency domain). In this paper, to build a zero-watermarking algorithm, the integer wavelet transform (IWT), is used alone to show the mathematical effect of the results obtained from the proposed zero-watermarking algorithms. One level of IWT is applied to the concealment image and the watermark is XORed with the features chosen from the crucial information obtained from the LL channel to generate the secret share. The results were very satisfactory and the proposed algorithm proved its resistance to different attacks through the robustness metric NC.

Pipelines are the safest as well as the most economical way to transport gas and condensate over long distances. Radiographic images are provided to commentators as a tool to diagnose welding defects in metal lines, so the study of welding in gas and oil pipelines has always been one of the most important areas of non-destructive testing. Expert interpreters are now used in many countries to interpret radiographic films from non-destructive tests. Interpreters can detect the number of pores on the weld surface by viewing radiographic images due to the limited number of these people and their unavailability. In some cases, there are many problems. For human interpretation, radiographic videos must be collected and sent to the interpreter's place of work or residence. The purpose of this article is to provide a method that can be used to interpret radiographic films quickly using conventional image processing methods and identify the welding defects in them and determine whether these defects are duplicates or not. The method of image segmentation is the area growth method. The main feature of this method is its proper performance in images such as radiographic images that have less subject variety. This method separates a part of the image from the rest by determining a pixel in the image as the starting point and expanding the area around this point due to the similarity between the pixels. In this paper, based on the histogram, the start and end image of the welding range is determined automatically. Then a combination of different standard algorithms is applied to identify defects in the image. Then, the key points of the image are extracted, and using them, the corresponding complex dynamic network is drawn and its calculations are performed. The simulation results show that the proposed method covers the shortcomings of the previous methods and in addition to bringing the detection of welding defects by computer closer to human diagnosis and in some cases works better than human performance, it has also made it possible to identify duplicate images.

In this paper, we use the mean curvature flow PDE and geodesic ODE to smooth and trace evolving curves as boundaries of minimal surfaces for a gray-scale image to capture their boundaries.

The reduction of the noise of the images always prevails as a challenge in the field of image processing. An image obtained after the elimination of noise has a higher clarity in terms of interpretation and study analysis in different fields such as medical, satellite and radar. This research work examines the various methods of de-noise images in the spatial domain and a comparison between several filtering techniques is carried out in the presence of different types of noise to achieve a high-quality image and to find the most suitable and reliable method for De-noising images. performance of all the filters is compared using parameters such as Mean Square Error (MSE), peak signal to noise ratio (PSNR).

Image processing by partial differential equations (PDEs) has been an active topic in the area of image denoising, which is an important task in computer vision. In PDE-based methods for unprocessed image process ing, the original image is considered as the initial value for the PDE and the solution of the equation is the outcome of the model. Despite the advan tages of using PDEs in image processing, designing and modeling different equations for various types of applications have always been a challenging and interesting problem. In this article, we aim to tackle this problem by introducing a fourth-order equation with flexible and trainable coefficients, and with the help of an optimal control problem, the coefficients are determined; therefore the proposed model adapts itself to each particular application. At the final stage, the image enhancement is performed on the noisy test image and the performance of our proposed method is compared to other PDE-based models.

Digital compression of images is a topic that has appeared in a lot of studies over the past decade to this day. As wavelet transform algorithms advance and procedures of quantization have helped to bypass current compression of image standards such as the JPEG algorithm. To get the highest effectiveness in compression of image transforms of wavelet need filters which gather a desirable character's number i.e., symmetry and orthogonally. Nevertheless, wave design capabilities are restricted due to their ability to have all of such desirable characters at the same time. The multi-wavelet technology removes a few of the restrictions of the wavelet play more than the options of design and thus able to gather all desired Characters of transforming. Wavelet and multi-wave filter banks are tested on a larger scale with images, providing more useful analysis. Multiple waves indicate energy-compression efficiency (a higher compression ratio usually indicates a higher mean square error, MSE, in the compressed image). Filter bank Characters such as orthogonal and compact support, symmetry, and phase response are important factors that also affect MSE and professed quality of the image. The current work analyzes the multi-wave Characters effect on the performance of compression of images. Four multi-wavelength Characters (GHM, CL, ORT4) were used in this thesis and the compression of image performance of grayscale images was compared with common scalar waves (D4). SPIHT quantification device in stress chart and use of PSNR and subjective quality measures to assess performance. The results in this paper point out those multi wave characteristics that are most important for the compression of images. Moreover, PSNR results and subjective quality show similar performance to the best scalar and multi-waves. The analysis also shows that a programmer based on multi-band conversion significantly improves the perceived image quality.

Face recognition has come to the top of the list of the most frequently used image processing applications, owing in large part to the availability of practical technology in this area. Despite significant progress in this sector, several issues such as ageing, partial blockage, and facial emotions impede the system's efficacy. Face identification from real-world data, recorded photographs, sensor images, and dataset images is difficult to solve because of the huge range of facial appearances, lighting effects, and complexity of the image background. Face recognition is a very successful and practical use of image processing and biometric systems. In this paper, we analyze the most significant challenges confronting the subject of face recognition; we discuss the challenges, how they were addressed using scientific methods, which databases are the most useful, and we summarize the most significant previous studies on age and gender that have been widely cited by researchers in the last year, along with a concise definition.

In this paper, we introduce an efficient conjugate gradient method for solving nonsmooth optimization problems by using the Moreau-Yosida regularization approach. The search directions generated by our proposed procedure satisfy the sufficient descent property, and more importantly, belong to a suitable trust region.  Our proposed method is globally convergent under mild assumptions. Our numerical comparative results on a collection of test problems show the efficiency and superiority of our proposed method. We have also examined the ability and the effectiveness of our approach for solving some real-world engineering problems from image processing field. The results confirm better performance of our method.

In this paper, a real-time denoising filter based on modelling of stable hybrid models is presented. The hybrid models are composed of the shearlet filter and the adaptive Wiener filter in different forms. The optimization of various models is accomplished by the genetic algorithm. Next, regarding the significant relationship between Optimal models and input images, changing the structure of Optimal models for image denoising is modelled by the ANFIS. The eight hundred digital images are used as train images. For eight hundred training images, Sixty seven models are found. For integrated evaluation, the amounts of image attributes such as Peak Signal to Noise Ratio, Signal to Noise Ratio, Structural Similarity Index, Mean Absolute Error and Image Quality Assessment are evaluated by the Fuzzy deduction system. Finally, for the features of a sample noisy image as test data, the proposed denoising model of ANFIS is compared with wavelet filter in 2 and 4 level , Fast bilateral filter, TV-L1, Median, shearlet filter and the adaptive Wiener filter. In addition, run time of proposed method are evaluated. Experiments show that the proposed method has better performance than others.

Annihilation or reduction of each kind of noise blended in correct data signals is a field that has attracted many researchers. It is a fact that fuzzy theory presents full capability in this field. Fuzzy filters are often strong in smoothing corrupted signals, whereas they have simple structures. In this paper, a new powerful yet simple fuzzy procedure is introduced for sharpness reduction in two-dimensional signals. It is indeed an extension of our previously published one-dimensional fuzzy smoothing filter. This procedure has been designed for annihilation of all unknown noises in two-dimensional corrupted signals, although works the best for impulse noise. The proposed method looks for emph{sharp points} in the corrupted signal and then smoothes them out by emph{sharing} their values with eight (or more) neighboring point values. Preservation of correct data in the corrupted signal is an important advantage of this method. To obtain experimental results of the proposed procedure, both color and black & white images are used as the most common two-dimensional signals, and the results are compared with several other filters recently cited in the literature. Experimental results exhibit a high capability of our method in both numerical measures and visual inspection, preserving its simplicity. Finally, application of the proposed filter to socio-economic fields is presented using a demographic mixed data set to better illustrate original motivation for this idea.