Zernike moments

In the paper we present an approach to Introduce automation of brain CT image analysis Because CT Scan method that used especially for the diagnosis of stroke and can detect bleeding in stroke due to a blocked artery, of course Images from a CT scan resolution is low relatively. Therefore, the grayscale images resolution is scant and makes detection difficult. We can use bioinformatics and artificial coloring techniques by image processing quality added and is more sensitive in outstanding. We have to identify and distinguish the areas of clustering artificial colors with Hopfield clustering that introduced as Pixel clustering based segmentation method and improve it by Hopfield neural network (HNN) based on spectral properties to show different region by artificial coloring and clustering. We want to improve the technique to use this rule by determining best cluster in neural network.

Floating-point arithmetics may lead to numerical errors when numbers involved in an algorithm vary strongly in their orders of magnitude. In the paper we study numerical stability of Zernike invariants computed via complex-valued integral images according to a constant-time technique from [2], suitable for object detection procedures. We indicate numerically fragile places in these computations and identify their cause, namely-binomial expansions. To reduce numerical errors we propose piecewise integral images and derive a numerically safer formula for Zernike moments. Apart from algorithmic details, we provide two object detection experiments. They confirm that the proposed approach improves accuracy of detectors based on Zernike invariants.

There are just too many trademarks out there so that a good automated retrieval system is required to help to protect them from possible infringements. However, it is from people, i.e., the general consumers' viewpoint how similar or confusing two trademarks can be. Thus, in this paper we propose a hybrid system where we patently incorporate human inputs into a computerized trademark retrieval scheme. Various surveys involving general consumers' cognition and responses are conducted, and the results are used as benchmarks in developing the automated part. The core mathematical features used in the scheme are fourgray-level Zernike moments and two new image compactness indices. Experimental results show that this hybrid system, when compared with humangenerated results, is able to achieve an average accuracy rate of 95% while that of the closest competing existing method is 65%.

In this paper a method for extraction of mid-level semantics from sign language videos is proposed, by employing high level domain knowledge. The semantics concern labeling of the depicted objects of the head and the right/left hand as well as the occlusion events, which are essential for interpretation and therefore for subsequent higher level semantic indexing. Initially the low-level skin-segement descriptors are extracted after face detection and color modeling. Then the respective labels are assigned to the segments. The occlusions between hands, head and hands and body and hands, can easily confuse extractors and thus lead to wrong interpretation. Therefore, a Bayesian network is employed to bridge in a probabilistic fashion the gap between the high level knowledge about the valid spatiotemporal configurations of the human body and the extractor. The approach is applied here in sign-language videos, but it can be generalized to any other situation where semantically rich information can be derived from gesture.

The importance of the plant for the human being and the environment led to deeply been studied and classified in detail. The advancement of the technology is the main factor in finding many ways for plant identification process. Some kind of initial intelligence systems in order to identify plant, followed by many theories and concepts using methods like; Moment Invariant (MI), Zernike Moments (ZM) and Polar Fourier Transform (PFT), and technologies for classification like Neural Network (NN), K-Nearest neighbour classifier (KNN) and Support Vector Machine (SVM), were used by many researchers through past years. In this paper is Centroid-Radii (C-R) combined with geometric features of the leaves, in order to cover most of the shape features of the leaves, colour moments, and the Grey-Level Co-occurrence Matrix (GLCM) to improve the accuracy of the system identification. In addition to the above features, Veins is also involved in the method being used, plus Principal Component Analysis (PCA), which is used to convert features into orthogonal features, and the results were inputted to the classifiers that used Probabilistic Neural Network (PNN). Two datasets have been used for testing: the first dataset is created especially for this work and collected from 24 kinds of plants, and the second dataset is called Flavia, which contains 32 kinds. The results were clearly improved to identify the plants. The maximum accuracy reached up to 98.50%, when using the first data set and 98.16% for the second dataset.

Most of Content Based Image Retrieval (CBIR) system use global texture features for representing and retrieving images. If local texture features are ignored during the initial stage of image processing, the performance will be affected. Meanwhile the features extraction, if it is based on Color co-occurrence Matrix (CCM) will provide the opportunities effective CBIR. Therefore, the main objective of this paper avoids the performance ineffectiveness and the same time opting for much effective CBIR. The problems that were highlighted will be tacked by considering the approach that is based on the local Haralick’s texture features, specifically using Average Analysis (AA) and Principal Component Analysis (PCA) methods. The extraction by Haralick’s texture feature was based on the predetermined CCM. The experimentation was done on the suggested ten categories of 1000 selected images from the Coral image database. The results portray, it is interesting to note that for certain image cat...

Today, the number of registered trademarks is huge and is increasing rapidly. Thus, the job of identifying infringement of trademarks by solely using manual inspection is tiring, laborious and time consuming. To cope with the tremendous amount of available registered trademarks and to protect the infringement of trademarks, a new automatic and efficient trademark retrieval system is necessary and urgent. This paper presents an efficient content-based trademark retrieval based on the synergy between the colorspatial technique and zernike moment method. Zernike moments could be used as an effective descriptor of global shape of a trademark while the color-spatial feature could be used to obtain color spatial distribution in the trademark image. To retrieve visually similarly look trademark, we use a weight Euclidean distance measure. Experimental evaluations conducted on a database containing 1000 trademark images shows that the proposed methodology is very effective in retrieving visually similar look trademarks.

The matching and retrieval of the 2D shapes are challenging issues in object recognition and computer vision. The principal idea in this paper focus to propose an robust descriptor for indexing and retrieval Arabic hand print character. If we make an observation, we can show that some Arabic letters are completely similar with small deferent is represented by a dots. In this paper, we propose new hybrid descriptor based on Fourier coefficients. The proposed vector descriptor is constructed with Fourier coefficients and tow scalars parameters, the first parameter given by Euler number and the second represent the number of objects in the shape on the binary image. The choice of Fourier descriptor as part in the combination of proposed descriptor is based on studies of four well known descriptors based on, Zernike moments, Invariants Hu moments, Elliptic Fourier coefficients and Fourier coefficients. The comparative studies show that the hybrid descriptor gives good results using a test database constructed by Arabic Hand print letters. The results given by the proposed descriptor show the perfection and robustness of our method.

This paper represents a new approach for face recognition that incorporates Prewitt edge detection, Gabor filter and Zernike moments to transform the image into a unified domain. On this joint domain, five distance metrics are constructed using Schoenberg transform for the purpose of defining efficient similarity measures for holistic face recognition. The proposed Schoenberg similarity applies Schoenberg transform to the logarithm of five existing distance metrics: Minkowski, City-Block, Euclidean, Soergel and Lorentzian metrics. The constructed Schoenberg logarithmic metrics are called SL-Minkowski, SL-City-Block, SL-Euclidean, SL-Soergel and SL-Lorentzian. These distance metrics are utilized as similarity measures after being normalized over the range [0,1] for fair comparison with existing measures. The proposed Schoenberg system can resist three problems: Change in illumination, pose and facial expression. Simulation results show that the proposed distance measures have superio...

This paper reports the results of experiments in improving performance of leaf identification system using Principal Component Analysis (PCA). The system involved combination of features derived from shape, vein, color, and texture of leaf. PCA was incorporated to the identification system to convert the features into orthogonal features and then the results were inputted to the classifier that used Probabilistic Neural Network (PNN). This approach has been tested on two datasets, Foliage and Flavia, that contain various color leaves (foliage plants) and green leaves respectively. The results showed that PCA can increase the accuracy of the leaf identification system on both datasets.

There are situations where it is not possible to capture a large document with a given imaging media such as scanner or copying machine in a single stretch because of their inherent limitations. This results in capturing a large document in terms of split components of a ...

The study presents a plant recognition system that uses image and data processing techniques for recognition. A lot of research has been going on to identify plants by their leaves and one of the features that is used is the shape of the leaf but the accuracy is not high and therefore other features should also be considered to increase the accuracy. This system designed has three main steps which are image pre-processing, feature extraction and matching. Image pre-processing performs basic operations on the leaf image for segmentation which helps in making feature extraction easy. Seven (7) leaf features derived from geometric parameters of leaf shape were extracted from the pre-processed image and the simple principle of minimum Euclidean distance was used for finding the closest match to the input leaf image. The system used 10 species of leaves with a total of 50 leaf images from the flavia dataset for testing and obtained an accuracy above 90%. The algorithm is accurate and is easy to implement. However, it is slow and not tested on a large dataset. It is hoped that this proposed system will be exploited further and the speed will be improved and will also be able to give more information on the plant.

The study presents a plant recognition system that uses image and data processing techniques for recognition. A lot of research has been going on to identify plants by their leaves and one of the features that is used is the shape of the leaf but the accuracy is not high and therefore other features should also be considered to increase the accuracy. This system designed has three main steps which are image pre-processing, feature extraction and matching. Image pre-processing performs basic operations on the leaf image for segmentation which helps in making feature extraction easy. Seven (7) leaf features derived from geometric parameters of leaf shape were extracted from the pre-processed image and the simple principle of minimum Euclidean distance was used for finding the closest match to the input leaf image. The system used 10 species of leaves with a total of 50 leaf images from the flavia dataset for testing and obtained an accuracy above 90%. The algorithm is accurate and is easy to implement. However, it is slow and not tested on a large dataset. It is hoped that this proposed system will be exploited further and the speed will be improved and will also be able to give more information on the plant.

Background and Objective: According to the World Health Organization, breast cancer is the main cause of cancer death among adult women in the world. Although breast cancer occurs indiscriminately in countries with several degrees of social and economic development, among developing and underdevelopment countries mortality rates are still high due to low availability of early detection technologies. From the clinical point of view, mammography is still the most effective diagnostic technology, given the wide diffusion of the use and interpretation of these images. Methods: Herein this work we propose a method to detect and classify mammographic lesions using the regions of interest of images. Our proposal consists in decomposing each image using multi-resolution wavelets. Zernike moments are extracted from each wavelet component. Using this approach, we can combine both texture and shape features, which can be applied both to the detection and classification of mammary lesions. We used 355 images of fatty breast tissue of IRMA database, with 233 normal instances (no lesion), 72 benign, and 83 malignant cases. Results: Classification was performed by using SVM and ELM networks with modified kernels in order to optimize accuracy rates, reaching 94.11%. Considering both accuracy rates and training times, we defined the ration between average percentage accuracy and average training time in a reverse order. Our proposal was 50 times higher than the ratio obtained using state-of-the-art approaches. Conclusions: As our proposed model can combine high accuracy rate with low learning time, whenever a new data is received, our work will be able to save a lot of time, hours, in learning process in relation to the best method of the state-of-the-art.

Background and Objective: Every second, on average, 8 (eight) new malware are created. So, our goal is to propose an antivirus, endowed with artificial intelligence, able of identifying malwares through models based on fast training and high-performance neural networks. Methods: Our NGAV (Next Generation Antivirus) is equipped with an authorial ELM (Extreme Learning Morphological) machine. Our bmELMs (Bitwise-Morphological ELMs) are inspired by the image processing theory of Mathematical Morphology. We claim that bmELMs are able to adapt in any machine learning dataset. Inspired by Mathematical Morphology, our bmELMs are capable of modeling any form present at the decisions boundaries of neural networks. Results: Our bmELMs results are compared with classical ELMs and evaluated through widely used classification metrics. Our antivirus, provided with Bitwise-Morphology, achieves an average accuracy of 97.88%, 93.07%, 93.07% and 91.74% in malware detection of PE (Portable Executable), Java, JavaScript and PHP, respectively. Conclusions: Our NGAV enables high performance, large capacity of parallelism, and simple, low-power architecture with low power consumption. We concluded that our Bitwise-Morphology assists to the main requirements for the proper operation and confection of antivirus in hardware.

This paper represents a new approach for face recognition that incorporates Prewitt edge detection, Gabor filter and Zernike moments to transform the image into a unified domain. On this joint domain, five distance metrics are constructed using Schoenberg transform for the purpose of defining efficient similarity measures for holistic face recognition. The proposed Schoenberg similarity applies Schoenberg transform to the logarithm of five existing distance metrics: Minkowski, City-Block, Euclidean, Soergel and Lorentzian metrics. The constructed Schoenberg logarithmic metrics are called SL-Minkowski, SL-City-Block, SL-Euclidean, SL-Soergel and SL-Lorentzian. These distance metrics are utilized as similarity measures after being normalized over the range [0,1] for fair comparison with existing measures. The proposed Schoenberg system can resist three problems: Change in illumination, pose and facial expression. Simulation results show that the proposed distance measures have superio...

Handwritten signature is broadly utilized as personal verification in financial institutions ensures the necessity for a robust automatic signature verification tool. This tool aims to reduce fraud in all related financial transactions' sectors. This paper proposes an online, robust, and automatic signature verification technique using the recent advances in image processing and machine learning. Once the image of a handwritten signature for a customer is captured, several pre-processing steps are performed on it including filtration and detection of the signature edges. Afterwards, a feature extraction process is applied on the image to extract Speeded up Robust Features (SURF) and Scale-Invariant Feature Transform (SIFT) features. Finally, a verification process is developed and applied to compare the extracted image features with those stored in the database for the specified customer. Results indicate high accuracy, simplicity, and rapidity of the developed technique, which are the main criteria to judge a signature verification tool in banking and other financial institutions.

One of important components in an image retrieval system is selecting a distance measure to compute rank between two objects. In this paper, several distance measures were researched to implement a foliage plant retrieval system. Sixty kinds of foliage plants with various leaf color and shape were used to test the performance of 7 different kinds of distance measures: city block distance, Euclidean distance, Canberra distance, Bray-Curtis distance,  2 statistics, Jensen Shannon divergence and Kullback Leibler divergence. The results show that city block and Euclidean distance measures gave the best performance among the others.

For finger-vein recognition, many successful methods, such as Line Tracking (LT), Maximum Curvature (MC) and Wide Line Detector (WL), have been proposed. Among these, LT has a very slow matching and featureextraction phase, and LT, MC and WL are translation and rotation dependent. Moreover, we show in the paper, they are affected by noise. To overcome these drawbacks, we propose using popular feature descriptors widely used for several Computer Vision or Pattern Recognition (CVPR) problems in the literature. The CVPR descriptors we test include Histogram of Oriented Gradients (HOG), Fourier Descriptors (FD), Zernike Moments (ZM), Local Binary Patterns (LBP) and Global Binary Patterns (GBP), which have not been applied to the finger-vein recognition problem before. We compare these descriptors against LT, MC, and WL and evaluate their running times, performance and resilience against noise, rotation and translation. We report that the LT and WL methods accuracy are comparable to each other and WL gives the best accuracy, LT method's speed is the slowest. Our results indicate that WL can be used together with ZM and GBP in case of rotation and noise, respectively.

For finger-vein recognition, many successful methods, such as Line Tracking (LT), Maximum Curvature (MC) and Wide Line Detector (WL), have been proposed. Among these, LT has a very slow matching and featureextraction phase, and LT, MC and WL are translation and rotation dependent. Moreover, we show in the paper, they are affected by noise. To overcome these drawbacks, we propose using popular feature descriptors widely used for several Computer Vision or Pattern Recognition (CVPR) problems in the literature. The CVPR descriptors we test include Histogram of Oriented Gradients (HOG), Fourier Descriptors (FD), Zernike Moments (ZM), Local Binary Patterns (LBP) and Global Binary Patterns (GBP), which have not been applied to the finger-vein recognition problem before. We compare these descriptors against LT, MC, and WL and evaluate their running times, performance and resilience against noise, rotation and translation. We report that the LT and WL methods accuracy are comparable to each other and WL gives the best accuracy, LT method's speed is the slowest. Our results indicate that WL can be used together with ZM and GBP in case of rotation and noise, respectively.

Floating-point arithmetics may lead to numerical errors when numbers involved in an algorithm vary strongly in their orders of magnitude. In the paper we study numerical stability of Zernike invariants computed via complex-valued integral images according to a constant-time technique from [2], suitable for object detection procedures. We indicate numerically fragile places in these computations and identify their cause, namely-binomial expansions. To reduce numerical errors we propose piecewise integral images and derive a numerically safer formula for Zernike moments. Apart from algorithmic details, we provide two object detection experiments. They confirm that the proposed approach improves accuracy of detectors based on Zernike invariants.

This paper describes an approach based on Zernike moments and Delaunay triangulation for localization of handwritten text in machine printed text documents. The Zernike moments of the image are first evaluated and we classify the text as handwritten using the nearest neighbor classifier. These features are independent of size, slant, orientation, translation and other variations in handwritten text. We then use Delaunay triangulation to reclassify the misclassified text regions. When imposing Delaunay triangulation on the centroid points of the connected components, we extract features based on the triangles and reclassify the text. We remove the noise components in the document as part of the preprocessing step so this method works well on noisy documents. The success rate of the method is found to be 86%. Also for specific handwritten elements such as signatures or similar text the accuracy is found to be even higher at 93%.

The last few years witnessed great and increasing use in the field of medical image analysis. These tools helped the Radiologists and Doctors to consult while making a particular diagnosis. In this study, we used the relationship between statistical measurements, computer vision, and medical images, along with a logistic regression model to extract breast cancer imaging features. These features were used to tell the difference between the shape of a mass (Fibroid vs. Fatty) by looking at the regions of interest (ROI) of the mass. The final fit of the logistic regression model showed that the most important variables that clearly affect breast cancer shape images are Skewness, Kurtosis, Center of mass, and Angle, with an AUCROC of 88% and an Accuracy of almost 89%. We also came to the conclusion that the Fibroid mass is small and less white than the Fatty mass

One of important components in an image retrieval system is selecting a distance measure to compute rank between two objects. In this paper, several distance measures were researched to implement a foliage plant retrieval system. Sixty kinds of foliage plants with various leaf color and shape were used to test the performance of 7 different kinds of distance measures: city block distance, Euclidean distance, Canberra distance, Bray-Curtis distance,  2 statistics, Jensen Shannon divergence and Kullback Leibler divergence. The results show that city block and Euclidean distance measures gave the best performance among the others.

Content based information retrieval is now a widely investigated issue that aims at allowing users of multimedia information systems to retrieve images coherent with a sample image. A way to achieve this goal is the automatic computation of features such as color, texture, shape, and position of objects within images, and the use of the features as query terms. In this paper we describe some results of a study on similarity evaluation in image retrieval using shape, texture, color and object orientation and relative position as content features. Images are retrieval based on similarity of features where features of the query specification are compared with features of the image database to determine which images match similar with the given features. Feature extraction is a crucial part for any of such retrieval systems.

One of important components in an image retrieval system is selecting a distance measure to compute rank between two objects. In this paper, several distance measures were researched to implement a foliage plant retrieval system. Sixty kinds of foliage plants with various leaf color and shape were used to test the performance of 7 different kinds of distance measures: city block distance, Euclidean distance, Canberra distance, Bray-Curtis distance,  2 statistics, Jensen Shannon divergence and Kullback Leibler divergence. The results show that city block and Euclidean distance measures gave the best performance among the others.

Image edge is the most basic feature of image. Most of the classical edge detection methods are sensitive to noise, poor anti-interference performance.To overcome the limitations of the earlier work a new approach has been proposed for color images using color gradients, hybrid approach which is ant colony optimization as well as canny (ANTC). Color based edge detection has 10 % more potential edges than the gray one. In practice, the more the adaptive parameters are provided, the more conveniently the proposed method can be used. In this paper, we employed two adaptive parameters like Peak signal to noise ratio and accuracy.

In the paper, an effective method for reducing Geometrical Error (G.E) and Numerical Error (N.E) of Zernike moments is proposed. By running MATLAB for the proposed Zernike's algorithm, the results of our proposed methods have shown a remarkable improvement to the total error of the analysis. The new proposed technique in reducing significantly Geometrical Error is performed better than that in the traditional technique. Considering two sides of G.E, N.E minimization and the reconstructed images having their size are almost with unchanged forms compared to the original images, then the proposed method has proven its potential capability in significantly reducing the two main errors of Zernike moments computation. Finally, the copy-move-rotate detection program has written by C++ under supporting OpenCV and Boost libraries that helps to verify the authentication of images.  Index Terms-tampered image detection, Zernike polynomial, geometric moments, region of interest Manuscript

One of important components in an image retrieval system is selecting a distance measure to compute rank between two objects. In this paper, several distance measures were researched to implement a foliage plant retrieval system. Sixty kinds of foliage plants with various leaf color and shape were used to test the performance of 7 different kinds of distance measures: city block distance, Euclidean distance, Canberra distance, Bray-Curtis distance,  2 statistics, Jensen Shannon divergence and Kullback Leibler divergence. The results show that city block and Euclidean distance measures gave the best performance among the others.

Leaf identification is a challenging research. So far, many approaches have been proposed. In this paper, an approach that combines Fourier descriptors with other shape features was investigated to identify 100 hundred kinds of leaves. The result shows that the combination of Fourier descriptors and several other shape features can be used to identify leaves with the accuracy rate of 88%. This result indicates that this approach is a promising way for identifying leaves.

The paper proposes a new approach as a combination of the multiscale analysis and the Zernike moment based for detecting tampered image with the formation of copy – move forgeries. Although the traditional Zernike moment based technique has proved its ability to detect image forgeries in block based method, it causes large computational cost. In order to overcome the weakness of the Zernike moment, a combination of multiscale and Zernike moments is applied. In this paper, the wavelets and curvelets are candidates for multiscale role in the proposed method. The Wavelet transform has successful in balancing the running time and precision while the precision of the algorithm applied the Curvelets does not meet the expectation. The comparison and evaluation of the performance between the Curvelets analysis and the Wavelets analysis combining with the Zernike moments in a block based forgery detection technique not only prove the effective of the combination of feature extraction and mul...

This paper presents a new method to classify facial expressions from frontal pose images. In our method, first Pseudo Zernike Moment Invariant (PZMI) was used to extract features from the global information of the images and then Radial Basis Function (RBF) Network was employed to classify the facial expressions, based on the features which had been extracted by PZMI. Also, the images were preprocessed to enhance their gray-level, which helps to increase the accuracy of classification. For JAFFE facial expression database, the achieved rate of classification in our experiment is 98.33%. This result leads to a conclusion that the proposed method can ensure a high accuracy rate of classification. Index Terms-Facial expression classification, pseudo Zernike moment invariant, RBF neural network.

Center Symmetric Local Binary Pattern (CSLBP) is widely used in texture and object detection, but its utilization in image hashing is still limited. Image hashing is a powerful technique to identify whether image content is changed or preserved. Two main issues in image hashing are its compact length and its robust discrimination power. To deal with these issues, a robust image hashing method is proposed by combining features of CSLBP and Discrete Wavelet Transform(DWT). CSLBP is applied on LL band of DWT for compact length hash. Discrimination power of hashing is enhanced by weigh factor which comprises information from all sub bands of 1-Level DWT. All sub-band images of 1-Level DWT is combined to form secondary image. Local statistical features like Standard Deviation and Magnitude Average are calculated to obtain local weight factor MASD (Magnitude Average and Standard Deviation). Weighted LL-CSLBP histogram is constructed using this weight factor. Experimental results are demonstrated with Receiver Operating Characteristics (ROC) and Normalized Hamming Distance (NHD). The results shows that, the proposed scheme is robust against content preserving operations and sensitive to malicious operations.

The rapid growth of digital imaging techniques and versatility among the images has motivated the development of Content Based Image Retrieval (CBIR). CBIR is the application of computer vision techniques which uses the visual contents to retrieve relevant images from large databases according to user's interests. The visual contents (color, texture, shape etc) serve as the features for the images. Features are measurements of extreme interest analysed from an image. This paper proposes a novel combined Color-Size and texture features for CBIR system. A new type of visual feature named Color-Size feature which integrates the information of both color and size of the image in terms of number of segments is proposed. To implement the system, initially the images are segmented using Watershed segmentation approach. Different images would yield different number of segments that has to be taken into account for the extraction of features. From the segmented image the Color-Size featu...

The purpose of this project was to do a comparative study for content based image retrieval using DCT, DWT, LBP and curvelet transform texture descriptor. The aim is to investigate an efficient and effective technique for content based image retrieval from DCT, DWT, LBP and discrete curvelet transform. Wavelet transform has performed better for one dimensional signal as it is good in representing point discontinuities. The discrete cosine transform (DCT) helps separate the image into parts (or spectral subbands) of differing importance (with respect to the image's visual quality).A local binary pattern (LBP) texture operator which labels the pixels of an image by thresholding the neighborhood of each pixel and considers the result as a binary number. In this work, different type of hundred images were collected and stored as the database. Then Color images were converted into gray color by using gray level cooccurrence matrix. After that it performs decomposition up to three levels by using discrete wavelet transform and For discrete cosine transform it takes cosine transform (returns real values of images) so we get frequency features. For local binary pattern which checks the difference in pixel values within a block. Here we got local features. For discrete curvelet transform wrapping based method was there and features are extracted. Then indexing, similarity matching was done by using Euclidian distance method. Hence finally project has been demonstrated by using the plotted graphs one for the efficiency and another graph with respect to time of the system. The results obtained demonstrate that the curvelet transform based retrieval are more efficient then DCT, DWT, LBP based retrieval. The results obtained are in accordance with the expected predictions of the existing theory of curvelet transform.

Signature recognition is the process of verifying a writer’s identity by checking the signature against samples previously stored in the database. Several techniques such as the distance-based and statistical classifiers used for feature extraction on a signature image are not invariant to scaling and rotation and the Scale invariant feature transform (SIFT) though invariant to scaling and rotation cannot cater for intra-class variation (Transposition) among set of genuine signature images. This paper proposed a SIFT-SURF algorithm which is used for enhanced offline signature recognition. The SIFT-SURF algorithm computes integral image; obtains hessian data and interest point for each computed integral image; applied neuro-scaling PCA based radial basis function neural network to compute the optimal features for each signature image to come up with an algorithm that is invariant to scaling and rotation as well as reliably match transposition among genuine samples of a signature imag...

This research focuses on image recognition of beef and pork. Beef as an example of halal food, while pork as haram food, especially for Muslims. This study used PNN classification and feature extraction methods. These images show some fundamental differences between pork and beef which based on colors and texture. Color was extracted by HSV model, otherwise texture extracted with 3 methods. These methods were Gabor, Principle Component Analysis (PCA) and Local Binary Pattern (LBP). Performance comparison of these methods was measured from the target accuracy of classification. Experiments conducted on 100 images of beef, pork and mixed, with attention to smoothing parameter (spread value/σ) in PNN and distribution data training and data testing. The best spread value obtained 10 for Gabor+HSV+PNN and LBP+HSV+PNN, but PCA+HSV+PNN was 108. The mixed meat was recognizable by PCA+HSV+PNN and LBP+HSV+PNN equal to 100%. The highest classification performance was achieved by PCA+HSV+PNN. T...

So far, plant identification has challenges for several researchers. Various methods and features have been proposed. However, there are still many approaches could be investigated to develop robust plant identification systems. This paper reports several experiments in using Zernike moments to build foliage plant identification systems. In this case, Zernike moments were combined with other features: geometric features, color moments and gray-level co-occurrence matrix (GLCM). To implement the identifications systems, two approaches has been investigated. First approach used a distance measure and the second used Probabilistic Neural Networks (PNN). The results show that Zernike Moments have a prospect as features in leaf identification systems when they are combined with other features.

Purpose: Breast cancer is among the leading causes of cancer death among women. The occurrence of breast cancer is similar both in developed countries and in underdeveloped and developing nations, although mortality is higher in underdeveloped countries due to late detection. Even though mammography is the most used technique for the differential diagnosis of breast cancer, breast thermography can be used as a complementary technique, more accurate than self-examination but accurate enough to guide the use of a mammogram. However, thermal imaging is still difficult for radiologists to understand. Machine learning can help improve this scenario. Deep Wavelet Neural Networks are convolutional neural networks that do not necessarily learn, as they may have predefined filter banks as their neurons. Methods: In this work, we propose a deep hybrid architecture to support breast thermography imaging diagnosis based on five-layer Deep-Wavelet Neural Networks, to extract attributes of region...

Weapon detection is a vital need in dual-energy X-ray luggage inspection systems at security of airport and strategic places. In this paper, a novel weapon detection framework in high- energy images of X-ray dual-energy images based on shape and edge features is proposed. In this framework, image enhancement is carried out by two noise removal and histogram stretching operations. Also, Connected Component Analysis (CCA) is applied to detect weapon candidate regions. An optimum feature set such as Fourier descriptors and invariant moments features are selected by feature forward selection algorithm and used to classify the detected objects into weapon (illicit) and non-weapon (lawful) objects using Probabilistic Neural Network (PNN) classifier. The proposed framework is evaluated on a perfect database consisting of various weapons in different size, type and view and accuracy rate of 96.48% has obtained.

One of important components in an image retrieval system is selecting a distance measure to compute rank between two objects. In this paper, several distance measures were researched to implement a foliage plant retrieval system. Sixty kinds of foliage plants with various leaf color and shape were used to test the performance of 7 different kinds of distance measures: city block distance, Euclidean distance, Canberra distance, Bray-Curtis distance,  2 statistics, Jensen Shannon divergence and Kullback Leibler divergence. The results show that city block and Euclidean distance measures gave the best performance among the others.

Recently, many approaches have been introduced by several researchers to identify plants. Now, applications of texture, shape, color and vein features are common practices. However, there are many possibilities of methods can be developed to improve the performance of such identification systems. Therefore, several experiments had been conducted in this research. As a result, a new novel approach by using combination of Gray-Level Co-occurrence Matrix, lacunarity and Shen features and a Bayesian classifier gives a better result compared to other plant identification systems. For comparison, this research used two kinds of several datasets that were usually used for testing the performance of each plant identification system. The results show that the system gives an accuracy rate of 97.19% when using the Flavia dataset and 95.00% when using the Foliage dataset and outperforms other approaches.

So far, plant identification has challenges for several researchers. Various methods and features have been proposed. However, there are still many approaches could be investigated to develop robust plant identification systems. This paper reports several experiments in using Zernike moments to build foliage plant identification systems. In this case, Zernike moments were combined with other features: geometric features, color moments and gray-level co-occurrence matrix (GLCM). To implement the identifications systems, two approaches has been investigated. First approach used a distance measure and the second used Probabilistic Neural Networks (PNN). The results show that Zernike Moments have a prospect as features in leaf identification systems when they are combined with other features.

Several researches in leaf identification did not include color information as features. The main reason is caused by a fact that they used green colored leaves as samples. However, for foliage plants-plants with colorful leaves, fancy patterns in their leaves, and interesting plants with unique shape-color and also texture could not be neglected. For example, Epipremnum pinnatum 'Aureum' and Epipremnum pinnatum 'Marble Queen' have similar patterns, same shape, but different colors. Combination of shape, color, texture features, and other attribute contained on the leaf is very useful in leaf identification. In this research, Polar Fourier Transform and three kinds of geometric features were used to represent shape features, color moments that consist of mean, standard deviation, skewness were used to represent color features, texture features are extracted from GLCMs, and vein features were added to improve performance of the identification system. The identification system uses Probabilistic Neural Network (PNN) as a classifier. The result shows that the system gives average accuracy of 93.0833% for 60 kinds of foliage plants.

Problem statement : One challenging research area nowadays is pattern recognition. Many applications lay under the field of pattern recognition such as face and iris recognition, speech recognition, texture discrimination and optical character recognition. A system that recognizes isolated pattern of interest is called pattern recognition. The pattern under consideration could be an image. During the process of image recognition, many problems could occur such as noise, distortion, overlap, errors in the segmentation results and obstruction of objects in the image. Several approaches to handle and solve pattern recognition problems have been developed. Examples of such approaches are neural networks (NN), Contour matching, texture and color signature. Approach : The aim of this study is to develop a system to recognize isolated fish object in the image based on a combination between significant extracted features using anchor points, texture and statistical measurements. A generic f...

The retrieval and classification of shape-based objects employing three descriptors-generic Fourier descriptor (GFD), Legendre moment descriptor (LMD), and wavelet Zernike moment descriptor (WZMD) are described. All three descriptors have been applied for shape retrieval to a database of 1000 shapes from 20 different classes, where each class consists of 50 shapes. The Euclidean distance has been calculated as a similarity measure parameter for shape classification. To study the effect of noise on the retrieval and classification, additive and multiplicative noise of various variances were applied to the database. The classification results have been compared and it is inferred that WZMD performs better than GFD and LMD techniques. Precision and recall were measured as parameters of performance metric. For retrieved shape recognition, an optical experiment employing joint transform correlator architecture has been carried out.

In most of classic plant identification methods a dichotomous or multi-access key is used to compare characteristics of leaves. Some questions about if the analyzed leaves are lobed, unlobed, simple or compound need to be answered to identify plants successfully. However, very little attention has been paid to make an automatic distinction of leaves using such features. In this paper we first explore if incorporating prior knowledge about leaves (categorizing between lobed simple leaves, and the unlobed simple ones) has an effect on the performance of six classification methods. According to the results of experiments with more than 1,900 images of leaves from Flavia data set, we found that it is statically significant the relationship between such categorization and the improvement of the performances of the classifiers tested. Therefore, we propose two novel methods to automatically differentiate between lobed simple leaves, and the unlobed simple ones. The proposals are invariant to rotation, and achieve correct prediction rates greater than 98%.

This paper represents a new approach for face recognition that incorporates Prewitt edge detection, Gabor filter and Zernike moments to transform the image into a unified domain. On this joint domain, five distance metrics are constructed using Schoenberg transform for the purpose of defining efficient similarity measures for holistic face recognition. The proposed Schoenberg similarity applies Schoenberg transform to the logarithm of five existing distance metrics: Minkowski, City-Block, Euclidean, Soergel and Lorentzian metrics. The constructed Schoenberg logarithmic metrics are called SL-Minkowski, SL-City-Block, SL-Euclidean, SL-Soergel and SL-Lorentzian. These distance metrics are utilized as similarity measures after being normalized over the range [0,1] for fair comparison with existing measures. The proposed Schoenberg system can resist three problems: Change in illumination, pose and facial expression. Simulation results show that the proposed distance measures have superio...