Pattern recognition (Engineering)

The positron emission tomography (PET) is a tool reference in routine clinical oncology. His applications has concerned the management and therapeutic monitoring, identification and definition of targets for radiotherapy, and quantification. The most frequently radiotracer's used are fluo-deoxyglucose(FDG) and Fluo-choline (18F-choline). In addition to the functional information, PET with algorithms of post treatment could give information on the stage of cell differentiation in HCC. The objective is to determine the stage of cell differentiation in HCC on imaging PET. We have apply the process of imaging post treatment on the image of PET Choline and PET FDG of patient whose stage of cell differentiation is known at histology. The same result as histology were found. The volume of HCC lesions detected using 18F-FDG was slightly better than that using 18F-choline. The capacity of these two radiotracers to detect rate of fixation of HCC depend of the type of lesion. This combinat...

The main objective of this project is cell nuclei segmentation and classification. In pre-processing convert the gray scale and apply the segmentation. In segmentation this project proposes the graph cut algorithm. After segmentation the features are extracts such as texture and etc. In classification this project used the feed forward neural network. The proposed method is very efficient. The feed forward neural network is very efficient for classification.

Fabric detection has an outstanding importance in inspection of fabric quality and defects. It works on principle of spectral estimation technique vision. it gets bonded to locate defected regions accurately. This project represented a most accepted method for patterned fabric defect detection and classification using spectral estimation technique and rough set theory. To extract the Regular pattern from the image of the patterned fabric, here use of Estimating Signal Parameter via Rotational Invariance Technique (ESPRIT) is done. In this technique the defected region i.e. the shape and location of the flawed areas are detected by comparing the pattern image and the source image also the rough set classifier is trained and tested to detect the types of defects in the patterned fabric image. Practically it is observed that this method can successfully be used to analyze & find the defects in patterned fabrics with nearly 96% success rate. This method is result oriented and better improving than previous method.

Garments conventionally address the protective, social and aesthetic needs of individuals but they can also be engineered to carry out a range of specific functions. Besides incorporating the features to meet the specific demands of a particular function, the garment must meet the basic requirements of protection and comfort. Pattern engineering for functional clothing involves applying technical, scientific and mathematical knowledge of patterns to modify and develop patterns with the objective of developing a garment which meets the function specific requirements, is comfortable to wear and contributes to maximize the efficiency and performance of the wearer. This paper explores, through examples, use of pattern engineering for functional garments to achieve these stated objectives. It establishes pattern engineering as the first step of planning a purposeful, efficient and aesthetic functional garment. It shows how the techniques of pattern engineering can be used to find solutions to challenges posed by the anthropometry of the human body and how these techniques are used to generate the blueprint of a functional garment incorporating all functional and aesthetic components.

Garments conventionally address the protective, social and aesthetic needs of individuals but they can also be engineered to carry out a range of specific functions. Besides incorporating the features to meet the specific demands of a particular function, the garment must meet the basic requirements of protection and comfort. Pattern engineering for functional clothing involves applying technical, scientific and mathematical knowledge of patterns to modify and develop patterns with the objective of developing a garment which meets the function specific requirements, is comfortable to wear and contributes to maximize the efficiency and performance of the wearer. This paper explores, through examples, use of pattern engineering for functional garments to achieve these stated objectives. It establishes pattern engineering as the first step of planning a purposeful, efficient and aesthetic functional garment. It shows how the techniques of pattern engineering can be used to find solutions to challenges posed by the anthropometry of the human body and how these techniques are used to generate the blueprint of a functional garment incorporating all functional and aesthetic components.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Image is a set of objects and extracting these object components from image has been a continuous area of research since very long time. In image object extraction the main emphasis is on extracting each or focused objects from the image. Image object extraction techniques first detect the objects in images and in second step extract the detected objects from the original image. The object extraction techniques commonly use methods like edge detection, segmentation and background subtraction to name a few. With the advancements of research in various other fields of science and technology over the period of time, these techniques have found their application in many different areas i.e. Medical disease detection, satellite object extraction, image retrieval, objects recognition etc. This paper gives brief

Problem statement: Heart localization is an important step in cardiac Magnetic Resonance Images (MRI) analysis. This study aims to locate the moving heart region from MRI sequence of images. Approach: The idea is to use the motion detection techniques to isolate the heart region from the background image and then apply morphological operations to construct a moving heart region mask. The mask is then applied to the MRI image to separate the Region Of Interest (ROI) that includes the heart. The K-means clustering algorithm is applied to the ROI to segment the heart walls. Results: Experimental results have shown that the performance of the proposed technique is superior to other MRI heart segmentation techniques in both complexity and accuracy. Conclusion: The proposed technique can be used as a pre segmentation step in any other future heart segmentation techniques to increase their accuracy through the localization of the moving heart region. The presented technique is fully automated technique and superior compared to other segmentation techniques.

The recent boom of artificial Neural Networks (NN) has shown that NN can provide viable solutions to a variety of problems. However, their complexity and the lack of efficient interpretation of NN architectures (commonly considered black box techniques) has adverse effects on the optimization of each NN architecture. One cannot simply use a generic topology and have the best performance in every application field, since the network topology is commonly fine-tuned to the problem/dataset in question. In this paper, we introduce a novel method of computationally assessing the complexity of the dataset. The NN is treated as an information channel, and thus information theory is used to estimate the optimal number of neurons for each layer, reducing the memory and computational load, while achieving the same, if not greater, accuracy. Experiments using common datasets confirm the theoretical findings, and the derived algorithm seems to improve the performance of the original architecture.

Pattern Recognition and Classification is considered one of the most promising applications in the scientific field of Artificial Neural Networks (ANN). However, regardless of the vast scientific advances in almost every aspect of the technology and mathematics, neural networks still need to be fairly large and complex (i.e., deep), in order to provide robust results. In this article, we propose a novel ANN architecture approach that aims to combine two fairly small Neural Networks based on an introduced probability term of correct classification. Additionally, we present a second ANN, used to reclassify the potentially incorrect results by using the most probable error-free results as additional training data with the predicted labels. The proposed method achieves a rapid decrease in the mean square error compared to other large and complex ANN architectures with a similar execution time. Our approach demonstrates increased effectiveness when applied to various databases, related t...

A lot of approaches, each following a different strategy, have been proposed in the literature to provide AdaBoost with cost-sensitive properties. In the first part of this series of two papers, we have presented these algorithms in a homogeneous notational framework, proposed a clustering scheme for them and performed a thorough theoretical analysis of those approaches with a fully theoretical foundation. The present paper, in order to complete our analysis, is focused on the empirical study of all the algorithms previously presented over a wide range of heterogeneous classification problems. The results of our experiments, confirming the theoretical conclusions, seem to reveal that the simplest approach, just based on cost-sensitive weight initialization, is the one showing the best and soundest results, despite having been recurrently overlooked in the literature.

Segmentation of brain tumor manually consumes more time and it is a challenging task. This paper detects the tumor inside the brain by doing segmentation and extraction of the tumor which is been detected. To prove the efficiency of the detection of brain tumor we have performed a comparative study of two segmentation algorithms namely “watershed segmentation algorithm” and “k-means clustering segmentation algorithm”. After the segmentation process the various morphological operations are applied on the segmented image. The morphological operations are applied to concentrate only on the required tumor part and ignoring the remaining area in the brain. The various thresholding algorithms like “Otsu’s thresholding” and “brute force thresholding” is applied to improve the efficiency of the final output image. Comparative study is made between the segmentation algorithms and the thresholding algorithms used. The further step of this project is to present an analytical method to detect t...

The new imaging techniques for organic body tissue give more details and information about the normal and abnormal tissue that help to distinguish the overlapping in margin of each organ. In this paper a new method of tumor line detection and segmentation is used to separate the abnormal from the normal surrounding tissue to get a real identification of involved and noninvolved area that help the surgeon to distinguish the involved area precisely. The method used in this paper is watershed method and image processing by using MatLab 7.1 to detect the tumor boundaries in MRI and CT image for different cases. Our result in this study is very clear for physician to distinguish the area of tumor for surgical planning.

Over the past decade, different image segmentation algorithms have been developed and employed in segmenting or analyzing brain magnetic resonance imaging (MRI) scans in the clinical applications for the detection of brain tumor. However, accurate detection, compression and transmission of brain tumor data remain parts of the challenging tasks militating against brain tumor telemedicine services due to the complex nature of brain tumor MRI scans. In overcoming this challenge, five different brain tumor segmentation algorithms were developed for this study. The algorithms were developed using MATLAB scripts. The developed algorithms were evaluated using patients' data retrieved from Mayfield website. The result of the comparative performance compression rate efficiency evaluation test carried out shows that the developed hybrid thresholdwatershed segmentation algorithm outperforms others in terms of compression efficiency. The result implies that the usage of the developed hybrid threshold-watershed segmentation algorithm in transmitting brain tumor patients' data transmission over wireless communication system will require limited bandwidth resource.

Image segmentation is a classic subject in the field of image processing and also is a hotspot and focus of image processing techniques. Many image segmentation ways are developed by the researchers so as to create pictures sleek and simple to judge. Since there is no general solution to the image segmentation problem, these techniques often have to be combined with domain knowledge in order to effectively solve an image segmentation problem. Till now, there is no perfect algorithm has been developed that could glance at all the factors present in image segmentation can be avoided. The aim of the researchers is to find out the issues of digital image segmentation that challenges the computer vision and enhance its accuracy and performance by planning new algorithms. In this paper, methodologies like K-Means clustering and watershed segmentation are discussed.

Remote sensing data is an important source of geological information. Several applications of different platforms were studied by many authors. One of the common applications was the geological mapping after applying various techniques of image processing. This paper aims to apply the fuzzy connectedness technique of image segmentation in solving the tasks of lithological characterization and trace various geologic contacts. This goal was achievable by using a new algorithm written by Matlab. The results after running the new code on the pilot area of G. Muwaylhah at Eastern Desert of Egypt were highly matches with those of the visual edge detection and mapping by running the classification algorism of ENVI commercial package.

The watershed segmentation methods are essential methods, to be considered for quick results in image handling and analysis. However, the main problem arises in produced image because it causes excess segmentation and noise. This research is conducted to improve this presented algorithm based on the mathematical morphology and filters to minimize flaws mentioned in that paper. Objective of this research is to find the gaps in the existing literary works. In most cases, themarker based segmentation is best because it marks the part of segment. The working of this proposed algorithm is checked by optimization of the part that is still an area of research.

The main objective of this project is fake currency detection using the image processing. Fake currency detection is a process of finding the forgery currency. After choose the image apply preprocessing. In pre-processing the image to be crop, smooth and adjust. Convert the image into gray color. After conversion apply the image segmentation. The features are extracting and reduce. Finally compare the image into original or forgery.

The three-dimensional (3D) modeling and recognition of 3D objects have been traditionally performed using local features to represent the underlying 3D surface. Extraction of features requires cropping of several local surface patches around detected keypoints. Although an important step, the extraction and representation of such local patches adds to the computational complexity of the algorithms. This paper proposes a novel Keypoints-based Surface Representation (KSR) technique. The proposed technique has the following two characteristics: (1) It does not rely on the computation of features on a small surface patch cropped around a detected keypoint. Rather, it exploits the geometrical relationship between the detected 3D keypoints for local surface representation. (2) KSR is computationally efficient, requiring only seconds to process 3D models with over 50000 points with a MATLAB implementation. Experimental results on the UWA and Stanford 3D models dataset suggest that it can accurately perform pairwise and multiview range image registration (3D modeling). KSR was also tested

Retinal vascular segmentation is the primary stage of any retinal image analysis system. This vascular segmentation is useful in automated retinal disease detection and in diagnosis system, as it is found that retina indicates early signs of various systemic diseases like hypertension, diabetes, cardiovascular diseases and retinopathy of prematurity. In this, retinal image segmentation plays a vital role in image processing; with this approach in this paper we have proposed an improved method for retinal blood vessel segmentation using morphological image processing. In this proposed work not only the entire retinal graph network is extracted from retina but here we classifies vessels branching and crossing in the whole network by assigning the class of end point, intersection point, splitting point. The results of given method are compared with all existing segmentation systems and here we found that our method outperforms by giving accuracy of 90% in vascular segmentation.

Watershed transform is a commonly used image segmentation method. The main problem with this segmentation technique is that of its sensitivity to noise and other irregularities which leads to over-segmentation. In this paper the over-segmentation problem is overcome by combing pre-processing and post-processing techniques along with watershed transform. First multi-scale morphological filtering by reconstruction is used to remove noise and then h minima transform is implemented to extract markers. These markers are then superimposed on gradient image. Watershed transform is then applied on the modified gradient map. Postprocessing region merging technique is used to merge the over segmented regions in the final segmented map. Experimental results show that the over-segmentation problem is reduced with the average segmentation accuracy of 0.96.

Detection and segmentation of Brain tumor accurately is a challenging task in MRI. The MRI image is an image that produces a high contrast images indicating regular and irregular tissues that help to distinguish the overlapping in margin of each limb. All automatic seed finding methods may suffer with the problem if there is no growth of tumor and any small white part is there. But when the edges of tumor is not sharped then the segmentation results are not accurate i.e. segmentation may be over or under. This may be happened due to initial stage of the tumors [5]. So , in this paper a method of tumor detection based on texture of the MRI and if it is detected then to segment it automatically is proposed in this paper to separate the irregular from the regular surrounding tissue to get a real identification of involved and noninvolved area that help the surgeon to distinguish the involved area precisely. The method used in this paper is texture analysis and seeded region growing met...

Monitoring crime on the highway is crucial process. Vehicle Detection and Tracking System from CCTV captured image for night vision. In night time, it is very difficult to identify the vehicles due to unwanted light sources. In this paper we present a method to eliminate the outer light sources by using gray scale image, dilation and Edge Detection. The images are taken at night time from the CCTV mounted on the highway

Biometric systems provide automatic identification of the people base on their own characteristic features. Unlike the other biometric systems such as face, voice, vein, fingerprint recognitions, iris has randomly scattered features. Iris recognition is considered as the one of the most reliable and accurate biometric identification system. It consists of four stages such as; image acquisition, image preprocessing, image feature extraction, and image matching process. In this work, we are proposing a new feature extraction method and new matching metric in order to find an effective threshold value to separate the intra and inter class distribution of the iris images of different people by using eightlevel quantization. Instead of using whole iris region, we have used statistically preselected iris regions. This selection had reduced the computational time and decreased the storage capacity. Our suggested metrics is rotation invariant and compares two vectors' selected rows calculated by the Fourier Transform. We have suggested eight level quantization of the phase information in order to create iris feature extraction. Finally, we have shown ROC Curves to check the accuracy of our proposed metric. The accuracy of our work is 99.0%, proposed threshold value is 0.746 where FAR is 0.07%, FRR is 39.32% and AUC is 0.97. Using the CASIA Iris Database, we have compared our proposed matching metric with Hamming Distance metric and we report better performance in terms of matching time. 

Security Design Patterns present the tried and tested design decisions made by security engineers within a well documented format. Patterns allow for complex security concepts, and mechanisms, to be expressed such that non domain experts can make use of them. Our research is concerned with the development of pattern languages for advanced crypto-systems. From our experience developing pattern languages we have encountered several recurring issues within security design pattern engineering. These issues, if not addressed, will affect the adoption of security design patterns. This paper describes these issues and discusses how they could be addressed.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bone is a nonlinear, inhomogeneous and anisotropic material. To predict the behavior of bones expert systems are employed to reduce the computational cost and to enhance the accuracy of simulations. In this study, an artificial neural network (ANN) was used for the prediction of displacement in long bones followed by ex-vivo experiments. Three hydrated third metacarpal bones (MC3) from 3 thoroughbred horses were used in the experiments. A set of strain gauges were distributed around the midshaft of the bones. These bones were then loaded in compression in an MTS machine. The recordings of strains, load, rate of loading, and displacement were used as ANN input parameters. The ANN which was trained using 3250 experimental data points from two bones predicted the displacement of the third bone (R 2 ≥ 0.98). It was suggested that the ANN should be trained using noisy data points. The proposed modification in the training algorithm makes the ANN very robust against noisy inputs measurements. The performance of the ANN was evaluated in response to changes in the number of input data points and then by assuming a lack of strain data. A finite element analysis (FEA) was conducted to replicate one cycle of force-displacement experimental data (to gain the same accuracy produced by the ANN). The comparison of FEA and ANN displacement predictions indicates that the ANN produced a satisfactory outcome within a couple of seconds, while FEA required more than 160 times as long to solve the same model (CPU time: 5 h and 30 min).

Remote sensing data is an important source of geological information. Several applications of different platforms were studied by many authors. One of the common applications was the geological mapping after applying various techniques of image processing. This paper aims to apply the fuzzy connectedness technique of image segmentation in solving the tasks of lithological characterization and trace various geologic contacts. This goal was achievable by using a new algorithm written by Matlab. The results after running the new code on the pilot area of G. Muwaylhah at Eastern Desert of Egypt were highly matches with those of the visual edge detection and mapping by running the classification algorism of ENVI commercial package.

Segmentation by watershed transform is a fast, robust and widely used in image processing and analysis, but it suffers from over-segmentation. We present in this paper some improvements to this algorithm based on the mathematical morphology in order to get over this difficulty. The performance of this method is validated on medical images. The results obtained show the good performance of this approach.

In classification problems, isotonic regression has been commonly used to map the prediction scores to posterior class probabilities. However, isotonic regression may suffer from overfitting, and the learned mapping is often discontinuous. Besides, current efforts mainly focus on the calibration of a single classifier. As different classifiers have different strengths, a combination of them can lead to better performance. In this paper, we propose a novel probability calibration approach for such an ensemble of classifiers. We first construct isotonic constraints on the desired probabilities based on soft voting of the classifiers. Manifold information is also incorporated to combat overfitting and ensure function smoothness. Computationally, the extended isotonic regression model can be learned efficiently by a novel optimization algorithm based on the alternating direction method of multipliers (ADMM). Experiments on a number of real-world data sets demonstrate that the proposed approach consistently outperforms independent classifiers and other combinations of the classifiers' probabilities in terms of the Brier score and AUC.

We make a comparison of five different face recognition algorithms. Four existed algorithm and a new Ensemble Voting Algorithm were be implemented. The algorithms training on a data set with 1815 faces, registering on a gallery set with 1027 faces, and ...

The Earth observation optical image is considered as the excellent applicant to extract the Earth's surface information. Cloud cover is a spectacular problem for land surface analysis. False detection and analysis typically results for even small percentages of cloud cover. Therefore knowing the percentage of cloud in the image is constructive for the user to allow for selection of those images that contain an acceptable amount of cloud cover (and therefore error) for their application. This paper presents the algorithm for cloud cover assessment using ROI (region of interest) image classification based on histogram segmentation for THEOS satellite images. The objective is to accurately determine the percentage of cloud cover in the optical satellite image, which is improved from the existing algorithm developed. In addition the morphology structure element is used in the algorithm. A study on the appropriate shape and size of structure element is required and has been discussed in this paper.

Proposed System detects Indian coins of different denomination. The spiraling business transaction at vending machines and automated systems working on token have spurred better coin recognition techniques, saddled with increased robustness. These techniques facilitate transaction making it easier in all forms of trade. Considering the essential factors, a system has been created which recognizes coin using digital image processing. The process performs notation check on the input image. The stated subsequent check enables the technique to endorse Rotation Invariance for capturing both side coin images. Thus comparison between the input object image and database image is performed. Further, plotting the resultant values gives minima which if less than a standard threshold establishes the recognition of the coin.

This paper provides survey on leaf disease detection technique by using image processing. India is an agricultural country and most of peoples wherein about 70% depends on agricultural. So leaf disease detection is very important research topic. Number of crops caused by fungi, bacteria etc. To overcomes this by using automatic leaf detection of plant by different image processing technique.

This paper provides survey on leaf disease detection technique by using image processing. India is an agricultural country and most of peoples wherein about 70% depends on agricultural. So leaf disease detection is very important research topic. Number of crops caused by fungi, bacteria etc. To overcomes this by using automatic leaf detection of plant by different image processing technique.

A brain tumor is an abnormal tissue growth in the brain in which cells grow and multiply uncontrollably. The main aim of our task is to identify brain tumor from a particular MRI image and to extract the tumor region. The proposed method uses various digital image processing techniques such as noise removal, Otsu's thresholding, morphological operations etc. This paper focuses on the use of marker-based watershed transform for segmentation, to get near about accurate results.

The point of this undertaking is to configuration, carry out and assess a picture handling programming based answer for programmed recognition and grouping of plant leaf infection. Anyway contemplates show that depending on unadulterated unaided eye perception of specialists to recognize and group infections can be tedious and costly, particularly in country regions and agricultural nations. So we present quick, programmed, modest and precise picture preparing based arrangement. Arrangement is made out of four primary stages; in the main stage we make a shading change structure for the RGB leaf picture and afterward, we apply shading space change for the shading change structure. Then, in the subsequent stage, the pictures are sectioned utilizing the K-implies bunching strategy. In the third stage, we figure the surface components for the portioned contaminated items. At long last, in the fourth stage the separated provisions are gone through a pre-prepared neural organization.

Image segmentation refers to the separation of objects from the background, and has been one of the most challenging aspects of digital image processing. Practically it is impossible to design a segmentation algorithm which has 100% accuracy, and therefore numerous segmentation techniques have been proposed in the literature, each with certain limitations. In this paper, a novel Falling-Ball algorithm is presented, which is a region-based segmentation algorithm, and an alternative to watershed transform (based on waterfall model). The proposed algorithm detects the catchment basins by assuming that a ball falling from hilly terrains will stop in a catchment basin. Once catchment basins are identified, the association of each pixel with one of the catchment basin is obtained using multicriterion fuzzy logic. Edges are constructed by dividing image into different catchment basins with the help of a membership function. Finally closed contour algorithm is applied to find closed regions...

In the previous works on heat stress detection, various state-of-the-art machine learning techniques had been used to detect stress patterns from electroencephalographic (EEG) signals. Since the handcrafted feature engineering-based approaches pose certain limitations and are sensitive to transform the nonlinearities of EEG, deep learning techniques have drawn attention in the domain of EEG-based applications. Moreover, existing approaches for stress detection consider the whole frequency band (delta to gamma) which conceals the redundant and lossy information that increased the false detection rate. Also, these approaches were implemented only to identify the stress in binary classes and confined their application to identifying the level of stress. Therefore, a multi-scale channel-wise convolution-based multi-level heat stress assessment has been proposed in this paper. The novel contributions of this work are (1) designing a multi-scale convolutional neural network (MS-CNN) for extracting precise information from individual frequency bands of EEG, (2) use of sparse connectivity to reduce the redundant information and increase the performance with less trainable parameters, and (3) multilevel heat stress assessment for precise identification of the severity of stress. The proposed approaches were evaluated on pre-recorded data of 10 rats in a simulated laboratory environment. The high accuracy of approximately 96–97% and 90–92% has been achieved for binary and multi-level stress detection. There is approximately a 6 to 50% reduction in the trainable parameters with a 2% improvement of accuracy achieved on adopting channel-wise convolution over MS-CNN and deep convolutional neural network (DCNN). This demonstrates the effectiveness of the adopted multiscale feature extraction and sparse connectivity to improve the performance with a less complex model.

The main objective of this project is fake currency detection using the image processing. Fake currency detection is a process of finding the forgery currency. After choose the image apply preprocessing. In pre-processing the image to be crop, smooth and adjust. Convert the image into gray color. After conversion apply the image segmentation. The features are extracting and reduce. Finally compare the image into original or forgery.

Imaging plays a central role in the diagnosis & treatment of brain tumor. Automated brain tumor detection from MRI Images is one of the most challenging tasks in today's Medical Imaging research. Magnetic Resonance Images are used to produce images of soft tissues of human body. Various segmentation techniques are proposed to find out tumored region in brain MRI images. In this paper marker based watershed transform, is used for segmentation before that some preprocessing techniques are used for noise removal and then with some simple morphological operations are found the tumor region with matlab.

Combining image segmentation based on the statistical classification with geometric prior information is supposed to increase robustness and reproducibility. A probability density function is initialized and a spatial constraint is defined which prevent segmentation that is not a part of the model. The goal of this work is a high quality image segmentation of healthy tissue and a precise delineation of tumor boundaries from multiple slices of MRI data. In this paper, algorithms like K-means, Watershed and Expectation Maximization algorithms are used for the investigation and the results of all the segmentations are compared. Based on the results, a common consensus on the robustness of each method is discussed. The Watershed segmentation and the Atlas are combined through markers and this has been applied in the Gray/White matter segmentation in MR images. A previous probability criterion is to be used for its calculation. These methods act as an aid in the early detection of many n...

The main objective of this project is fake currency detection using the image processing. Fake currency detection is a process of finding the forgery currency. After choose the image apply preprocessing. In pre-processing the image to be crop, smooth and adjust. Convert the image into gray color. After conversion apply the image segmentation. The features are extracting and reduce. Finally compare the image into original or forgery.

Image segmentation refers to the separation of objects from the background, and has been one of the most challenging aspects of digital image processing. Practically it is impossible to design a segmentation algorithm which has 100% accuracy, and therefore numerous segmentation techniques have been proposed in the literature, each with certain limitations. In this paper, a novel Falling-Ball algorithm is presented, which is a region-based segmentation algorithm, and an alternative to watershed transform (based on waterfall model). The proposed algorithm detects the catchment basins by assuming that a ball falling from hilly terrains will stop in a catchment basin. Once catchment basins are identified, the association of each pixel with one of the catchment basin is obtained using multicriterion fuzzy logic. Edges are constructed by dividing image into different catchment basins with the help of a membership function. Finally closed contour algorithm is applied to find closed regions...

Biometric systems provide automatic identification of the people base on their own characteristic features. Unlike the other biometric systems such as face, voice, vein, fingerprint recognitions, iris has randomly scattered features. Iris recognition is considered as the one of the most reliable and accurate biometric identification system. It consists of four stages such as; image acquisition, image preprocessing, image feature extraction, and image matching process. In this work, we are proposing a new feature extraction method and new matching metric in order to find an effective threshold value to separate the intra and inter class distribution of the iris images of different people by using eightlevel quantization. Instead of using whole iris region, we have used statistically preselected iris regions. This selection had reduced the computational time and decreased the storage capacity. Our suggested metrics is rotation invariant and compares two vectors' selected rows calculated by the Fourier Transform. We have suggested eight level quantization of the phase information in order to create iris feature extraction. Finally, we have shown ROC Curves to check the accuracy of our proposed metric. The accuracy of our work is 99.0%, proposed threshold value is 0.746 where FAR is 0.07%, FRR is 39.32% and AUC is 0.97. Using the CASIA Iris Database, we have compared our proposed matching metric with Hamming Distance metric and we report better performance in terms of matching time. 

In this paper we study the performance of Spiking Neural Networks (SNN)and Support Vector Machine (SVM) by using a GPU, model GeForce 6400M. Respect to applications of SNN, the methodology may be used for clustering, classification of databases, odor, speech and image recognition..In case of methodology SVM, is typically applied for clustering, regression and progression. According to particular characteristics of these methodologies,theycan be parallelizedin several grades. However, level of parallelism is limited to architecture of hardware. So, is very sure to get better results using other hardware with more computational resources. The different approaches are evaluated by the training speed and performance. On the other hand, some authors have coded algorithms SVM light, but nobody has programming QP SVM in a GPU. Algorithms were coded by authors in the hardware, like Nvidia card, FPGA or sequential circuits that depends on methodology used, to compare learning timewith between GPU and CPU. Also, in the survey we introduce a brief description of the types of ANN and its techniques of execution to be related with results of researching.

Imaging plays a central role in the diagnosis & treatment of brain tumor. Automated brain tumor detection from MRI Images is one of the most challenging tasks in today’s Medical Imaging research. Magnetic Resonance Images are used to produce images of soft tissues of human body. Various segmentation techniques are proposed to find out tumored region in brain MRI images. In this paper marker based watershed transform, is used for segmentation before that some preprocessing techniques are used for noise removal and then with some simple morphological operations are found the tumor

Implementing automated emotion recognition on mobile devices could provide an accessible diagnostic and therapeutic tool for those who struggle to recognize emotion, including children with developmental behavioral conditions such as autism. Although recent advances have been made in building more accurate emotion classifiers, existing models are too computationally expensive to be deployed on mobile devices. In this study, we optimized and profiled various machine learning models designed for inference on edge devices and were able to match previous state of the art results for emotion recognition on children. Our best model, a MobileNet-V2 network pre-trained on ImageNet, achieved 65.11% balanced accuracy and 64.19% F1-score on CAFE, while achieving a 45-millisecond inference latency on a Motorola Moto G6 phone. This balanced accuracy is only 1.79% less than the current state of the art for CAFE, which used a model that contains 26.62x more parameters and was unable to run on the ...