Papers by J.D.Dorathi Jayaseeli
International Journal on Advanced Science, Engineering and Information Technology, 2020
Image segmentation plays a crucial role in image analysis processes. The operations performed on ... more Image segmentation plays a crucial role in image analysis processes. The operations performed on a segmented image tend to affect it differently than if they were performed on the original image; therefore, segmenting an image can show radically different results from the original image and successfully do so can yield features and other important information about the image. Proper image analysis is of high importance to the medical community as accurately classifying different conditions and diseases can be facilitated with excellent patient imaging. Multifractal analysis can be leveraged for performing texture classification and image segmentation. In this paper, we propose fusion-based algorithms utilizing multifractal analysis for medical image segmentation. We use two specific multifractal masks: square and quincunx. Our techniques show new insights by using methods such as histogram decomposition in conjunction with new techniques, such as fusion. By fusing different slope images, we can extract more features, thus making our proposed algorithms more robust and accurate than traditional multifractal analysis techniques. These methods are further capable of reliably segmenting medical images by implementing multifractal analysis techniques in coordination with methods such as gaussian blurring and morphological operations. Medical professionals can easily analyze the resulting image for diagnosing medical conditions. The outcomes show that the proposed algorithms extract dominant features that are more encompassing and powerful than classical techniques.
2021 Third International Conference on Intelligent Communication Technologies and Virtual Mobile Networks (ICICV)
In the current world, more confidential information on untrusted repositories is maintained such ... more In the current world, more confidential information on untrusted repositories is maintained such that it is important to encrypt data on these pages. New encryption methods are being developed and improved from time to time. "AES" (Advanced Encryption Standard) is one such encryption protecting machine, which offers fine grained access control. Think of the ABE as a Cipher Scheme. However, the expressive mode of the ABE method can be problematic for the ABE in interpreting the key clarification and result in a low computational performance for the key problem and key decryption. To fix this issue, several outsourced versions of the Chinese Abacus (a concrete plane not made of beads on sticks), were built to improve the reliability of its use. This paper describes the various methods suggested and contrasts them, illustrating the benefits with the drawbacks.
This paper presents a fuzzy filtered neural network approach as an application to handwritten num... more This paper presents a fuzzy filtered neural network approach as an application to handwritten numerical representation. A multilayer feedforward adaptive network is used for training the model and for application of the fuzzy filters. Fuzzy filters are integrated with the neural nets for processing the physical data of the images available for the handwritten digits. The use of the fuzzy filters reduces the noise and redundancy present in the data which ultimately increases the performance of the model. This also helps in avoiding the high complexity of the neural network architecture which would otherwise be required for the same physical data. Different varieties of the fuzzy filters are integrated with the neural network separately and their performance is compared. The fuzzy filters with higher dimensionality improve the model recognition rate. One dimensional and two dimensional fuzzy filters are discussed and their performance is evaluated. Finally, genetic algorithm based fuz...
International Journal of Engineering and Technology, 2016
Human society is living in the age of high speed development and increase of high volume of visua... more Human society is living in the age of high speed development and increase of high volume of visual data. It has become a necessity to evaluate the quality of these visual data or images in many applications. This approach of quality assessment focuses on collecting feature vector from Natural Scene Images that reflects the shape parameter and correlation between adjacent pixels. A set of 36 features is collected from each image and training is done on images available in the data set. Then, three learning approaches-General Classification, KNN approach, and Distortion Specific approach are done on this feature vector. The 0training is done by taking the Mean Opinion Score (MOS) values available at the TID data set. Each method reflected its applicability and accuracy. It has been observed that distortion specific method out performs the other two. Specifically, the area where each one of these methods can be applied is also identified which is of great help to image manipulators. Keyword-Blind/Non-Reference, Distortion Specific, Normalized Luminance, Quality Assessment I. INTRODUCTION The age we live in is a highly digital world. Progresses in technology have helped in images and videos to be captured, stored, shared and viewed easily. All these types of manipulations happening to the image gradually lead to a reduction in quality or content in the image. This led to an increase in development of tools for assessing image quality. Image Quality Assessment (IQA) tools help in rating the way humans perceive the quality of images. It is now a challenge for research community to develop the most accurate quality assessment tool. Investigators in quality assessment of images have endeavored to explore how the existence of these distortions disturbs the viewing experience. The best measure of quality assessment is to exhibit the images to a large sample of human population and to collect their opinion. The average of those opinions will give the average quality measure commonly indicated by subjective assessment. In objective assessment, the evaluation of quality is done using algorithms. The subjective assessment of quality takes lot of time and is impractical. Current IQA methods can be categorized into three based on the presence of the original image: i) complete/full reference IQA [15]; ii) less/reduced reference IQA; and 3) without/no reference/Blind IQA [18] (BIQA). BIQA is applied more in real world. Human eyes can easily distinguish the distortions or noises in natural images. This is because of the fact that there exist particular structures that separate the unnatural from the natural scenes. Such structures are called Natural Scene Statistics (NSS) [11]. Also, natural images are highly non-random with interdependencies within them [19], [20]. Earlier work in BIQA was based on knowledge of the type of distortion happened to the image and then, as developments happened , the distortion detecting algorithm determined the type of distortion and based on that, quality was assessed [5],[7]. Recently, techniques have been established to directly plot features in images to quality scores without looking into the type of distortion [10]. For example, Saad et al. [5] has developed the methods BLIINDS and BLIINDS-II [12] to degenerate the image features to quality scores. Quality Assessment Based on DCT: Michele Saad, Alan C. Bovik, Christophe Charrier developed a method called by the name BLIINDS [5] to effectively assess quality in Natural Scene Statistics (NSS). The algorithm relies on Discrete Cosine Transform (DCT) for feature extraction. The local DCT contrast value is calculated and is used for determining the distortion. The histogram pattern of distorted images exhibits a higher peak at '0' and the variance value also gets transformed accordingly. The DCT coefficient Kurtosis value is computed for this. The degree of this peak and tail weight are quantified to get the distortion index. Anush Krishna Moorthy and Alan C. Bovik [4] in 2011 devised a method for blind image quality assessment using wavelet transform coefficients (DIIVINE). Here a loose wavelet transform is applied on to the image and the scale-space-orientation of the image is noted. They supply a set of statistical features and they are stacked to form a vector. Using this feature vector, the distortion type is determined and the quality score is calculated. Here, a regression model is developed for each distortion category and mapping is done from distortion to quality value..
Proceedings of International Conference on Deep Learning, Computing and Intelligence
This paper aims at applying the techniques of deep learning and study the behavior of its effecti... more This paper aims at applying the techniques of deep learning and study the behavior of its effective score comparing with traditional approaches like supervised learning and proposed to come up with a revised algorithm with application towards hand written character recognition using the principles of deep learning architecture and analyze its performance with the conjunction of benchmarking machine learning dataset like MNIST. Hand written character recognition is achieved using the deep learning model namely Deep Belief Network which is trained using a simple Restricted Boltzmann Machine (simple RBM) and three layers of Restricted Boltzmann Machine (stacked RBM). The performance of our model shows 92% accuracy. This shows that it outperforms the traditional supervised learning methods. This method can be extended to efficient text extraction in complex images.
The paper describes in detail the techniques used for edge detection in images as well as evoluti... more The paper describes in detail the techniques used for edge detection in images as well as evolutionary algorithms which can be applied for Multi-level thresholding. In the first section it describes the core idea of edge detection followed by the first derivative operators being used currently for edge detection like: Roberts,Sobel,Prewitt,Robinson and Kirsch. The next section continues with explanation regarding the second derivative operators like: Laplacian, of Gaussian, Difference of Gaussians and Marr-Hildreth Algorithm. Further,it is followed by techniques used to detect edges using region based method known as Multi-Level Thresholding which is being applied using either the Otsu or the Kapur method. Evolutionary Algorithms namely Particle Swarm optimisation, Firefly Optimisation Algorithm and Elephant Herding Algorithm are explained in context with Kapur's method for Multi-Level Thresholding. Finally, experimental results regarding one common image and application of the ...
2021 IEEE 4th International Conference on Computing, Power and Communication Technologies (GUCON), 2021
With the advantage of the web and present-day innovation, it has been made conceivable to direct ... more With the advantage of the web and present-day innovation, it has been made conceivable to direct and conduct daily classes and regular assessments for students at educational institutions remotely. This has become all the more important with the recent shift of working life for students and teachers to the online realm. Nonetheless, this sort of web-based learning does not have the same advantage of intuitive human interaction and correspondence that physical homeroom learning has. In order to improve the experience of web-based learning, instructors may find that it is useful to have some system to alert them when a student appears to stop focusing during online sessions. Facial recognition and facial expression recognition have seen great strides forward in the recent years, with various methods being developed to facilitate the detection of human faces as well as the classification of facial expressions. Various techniques have seen success in the past, such as the use of machine learning techniques that use Histogram of Oriented Gradients (HOG), Local Binary Patterns (LBP), etc., and deep learning methods such as Residual Neural Network (RESNET), Visual Group Geometry 16 (VGG16), LeNet, to name but a few. To solve this problem of waning student concentration due to lack of human supervision during online classes, this paper aims to develop a Students Attention Monitoring and Alert Model (S-AMAM).
Intelligent Computing and Innovation on Data Science, 2021
Detecting the damage on a car is an image-based processing method with enormous scope for automat... more Detecting the damage on a car is an image-based processing method with enormous scope for automation. This concept of automated detection of the extent of exterior damage on a car and subsequent quantification of the damage severity would benefit car insurers, car rentals and repair services. In this paper, we propose employing convolution neural networks to build a Mask R-CNN model that can detect the area of damage on a car. The dataset used consists of images of damaged vehicles with a single class named scratch. The images are precisely annotated with the area of damage. The model is trained using the base weights of Mask R-CNN COCO dataset. The images are processed for 21 epochs. After processing, the final result is visualized using a color splash technique, wherein the area of damage is highlighted. This model would help in reducing the cost of processing insurance claims and lead to greater customer satisfaction. Car dealers can eliminate the manual process of damage assessment and the labor cost accompanied by it. Accuracy and transparency in pricing cars and their potential repairs will be made more prevalent. Fraudulent vehicle insurance claims can also be diminished. On implementing our model, we achieved an overall loss of 0.3888.
Procedia Computer Science, 2020
In this paper, an improved cuckoo search optimization algorithm for extracting road regions from ... more In this paper, an improved cuckoo search optimization algorithm for extracting road regions from high resolution images using multi-level thresholding schema is presented. Automatic road region extraction from high resolution satellite images automatically neglects sharpening the image segments since the available information is with high pixel values. However, occlusion and overlapping of objects are yet another challenging task in segmenting the roads from available images. And also identifying the number of threshold values which defines all type of roads (main roads and roads alongside main roads, etc.) increases the complexity of the problem to define exact road region. In this proposed method, multi-level thresholding concept is applied for efficient road region extraction (Otsu). After finding the number of available threshold values to be segmented, an improved cuckoo search optimization algorithm is incorporated for finding the optimal threshold value for extracting the road regions from the given image. The conventional multiclass SVM classifier is used for efficient extraction of road regions from the given images. This proposed methodology will be tested with three developed suburban region satellite images and the results are compared with existing segmentation algorithms.
Journal of Physics: Conference Series, 2018
View the article online for updates and enhancements. You may also like Can arousing feedback rec... more View the article online for updates and enhancements. You may also like Can arousing feedback rectify lapses in driving? Prediction from EEG power spectra Chin-Teng Lin, Kuan-Chih Huang, Chun-Hsiang Chuang et al. Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
International Journal of Engineering & Technology, 2018
Extraction of Roads, Rivers and other map objects is an important step in many military and civil... more Extraction of Roads, Rivers and other map objects is an important step in many military and civilian applications. In this process the information is extracted which possess high efficiency and accuracy and is fed into GIS (Geographical Information System). In this paper, we have explored different algorithms with better efficiency and accuracy. Road extraction can take place for two kinds of roads namely: urban and non-urban roads. Urban roads are more complex to analyze because of their architectural complexity, occlusions created by trees, heavy traffic and extensive network, whereas non-urban roads are easier to analyze because of less structural complexity. The proposed algorithm exploits the properties of road segments to develop customized operators to accurately derive the road segments. The customized operators include directional morphological enhancement, directional segmentation and thinning. The proposed algorithm is systematically evaluated on the basis of v...
International Journal of Engineering & Technology, 2018
The text-based password has been the most common practice from ancient days till present. Text ba... more The text-based password has been the most common practice from ancient days till present. Text based pass-words are also known for various threats, and it is prone to attacks like guessing attacks, dictionary attacks, social engineering attacks, brute force attacks, etc. The next immediate concept following the text based password is the graphical password schemes to improve password security and usability. In present days graphical passwords are being implemented more commonly. This approach is different from the traditional alpha numeric as it deals with images. In this paper a survey study is done to analyse various techniques used for authentication and also some of the methods for graphical authentication techniques like Pass Matrix, Cued Clicked points(CPP), CAPTCHA, Image distortion with text association, Doodle scheme, Standard recognition-based scheme, Stegno pin authentication method. Based on the existing methods, the future research can be done in order to improve securi...
International Journal of Engineering & Technology, 2018
The invention of the net has introduced the unthinkable growth and developments within the illust... more The invention of the net has introduced the unthinkable growth and developments within the illustrious analysis fields like drugs, satellite imaging, image process, security, biometrics, and genetic science. The algorithms enforced within the twenty first century has created the human life more leisurely and secure, however the protection to the first documents belongs to the genuine person is remained as involved within the digital image process domain. a replacement study is planned during this analysis paper to discover. The key plan in the deliberate take a look at and therefore the detection of the suspected regions are detected via the adaptive non-overlapping and abnormal blocks and this method is allotted exploitation the adaptive over-segmentation algorithmic rule. The extraction of the feature points is performed by playacting the matching between every block and its options. The feature points are step by step replaced by exploitation the super pixels within the planned F...
International Journal of Engineering & Technology, 2018
The Road extraction from aerial image, stands as a quintessential node for the development of rud... more The Road extraction from aerial image, stands as a quintessential node for the development of rudimentary layers in innumerable fields. From GIS, to Unmanned Aerial vehicles, road maps pave the foundation for data accumulation. This significant process is a result of number of mechanisms devised over the years through iterative experiments and research. However, the glut of methods available often pose as a hurdle in the selection process. In this project we implement a novel approach to solve the extraction problem, by incorporating generative algorithm using conditional adversarial networks. We investigate conditional adversarial networks as a general-purpose solution to image-to-image translation problems. These networks not only learn the mapping from input image to output image, but also learn a loss function to train this mapping. This makes it possible to apply the same generic approach to problems that traditionally would require very different loss formulations. The U-Netw...
Uploads
Papers by J.D.Dorathi Jayaseeli