Search

With the advancement in technology, Computer and machine vision system is getting involved in the agriculture sector for the last few years. Deep Learning is a recent advancement in the Artificial Intelligence field. In the present era, many researchers have used deep learning applications for the classification of images, and is found to be one of the emerging areas in computer vision. In the classification of fruit images, the main goal is to improve the accuracy of the classification system. The accuracy of the classifier depends on various factors like the nature of acquired images, the number of features, types of features, selection of optimal features from extracted features, and type of classifiers used. In the proposed article, integration of CNN, RNN, and LSTM for the classification of fruit images are defined. In this approach, CNN and RNN are employed for the development of discriminative characteristics and sequential-labels respectively. LSTM presents an explanation by integrating a memory cell to encode learning at each interval of classification. Key parameters: accuracy, F-measure, sensitivity, and specificity are applied to assess the achievement of the proposed scheme. From empirical results, it has been declared that the offered classification method provides efficient results.

Fruit classification is noticed as the one of the looming sectors in computer vision and image classification. A fruit classification may be adopted in the fruit market for consumers to determine the variety and grading of fruits. Fruit quality is a prerequisite property from health view position. Classification systems described so far are not adequate for fruit classification during accuracy and quantitative analysis. Thus, the examination of new proposals for fruit classification is worthwhile. In the present time, automatic fruit classification is though a demanding task.Deep learning is a powerful state of the art approach for image classification [1] This task incorporates deep learning models: Convolution Neural Network (CNN), Recurrent Neural Network (RNN) and Long Short Term Memory (LSTM) for classification of fruits based on chosen optimal and derived features. As preliminary arises, it has been recognized that the recommended procedure has effective accuracy and quantitative analysis results. Moreover, the comparatively high computational momentum of the proposed scheme will promote in the future for the real time classification operations

Artificial Intelligence, Machine Learning, Fuzzy Logic, Neural Network, Genetic Algorithm and their hybrid systems play vital role in the medical sciences to diagnose various diseases efficiently in the patients. The problems related to the heart are widely comon in today’s world. The risk of heart failure develops due to the narrowness and blockage in the coronary arteries of the heart as excess cholesterol deposits in the arteries and blood vessels that results in fatigue, chest pain, dyspnoea, sleeping difficulties and depression. This research aims to explore diverse work done on FL and Hybrid-based techniques to identify the risk of heart disease among the patients. The present study reveals publications along with the strength, operating system, accuracy rate and other specifications used in the identification of heart disease based on FL and Hybrid-based approaches since 2010. This survey contributes motivation for research scholars to generate more innovative ideas and continue their research work in the respective field. Moreover, the future model for direct service of the patients from old age homes to the Intensive Care Unit through ambulance services is also presented in this paper.

Image Enhancement is used as a preprocessing step in many computer vision applications. It provides enhanced input for other computerized image processing methods. Many preprocessing techniques can be applied to images depending on the application domain. In this paper we are proposing an image enhancement technique for color images that can be used as preprocessing step in many computer vision applications. It can also be used as a data augmentation technique in object detection. Luminance component of images is sometimes not captured by cameras and displayed by monitors properly. To remove this drawback of devices we have used gamma encoding. Four different values of gamma are evaluated depending on the quality of images. Image is then converted into YUV Color space. Y component represents the luminance. U and V components represent color. After that Contrast Limited Adaptive Histogram Equalization is applied to the Y component to improve the contrast of the image. The results are compared with the state-of-the-art methods on the basis of Peak Signal to noise Ratio (PSNR) and Mean Square Error (MSE). Quantitative results show that proposed algorithm results in improved value of PSNR and decreased value of MSE as compared to existing methods. Qualitative comparison is also done and results show improvement over the existing techniques.

We suggested a Fuzzy image enhancement technique using Type-II to improvise visibly weather-degraded fruit images. The acquired fruit picture is divided into dark and light regions with Fuzzy partitioned. Fuzzy-sure entropy was used to choose the optimum visibility enhancement threshold values. Compared to the ATCE, BPDHE, GC and CLAHE image enhancement frameworks, the proposed scheme offers improved image quality. Experimental findings show that the approach in terms of visual and quantitative analysis have a lower percentage of saturated pixels, better output of visible edges and color gradients.

Deep learning has been a game changer in the field of object detection in the last decade. But all the deep learning models for computer vision depend upon large amount of data for consistent results. For real life problems especially for medical imaging, availability of enough amounts of data is not always possible. Data augmentation is a collection of techniques that can be used to extend the dataset size and improve the quality of images in the dataset by a required amount. Logically it is used to make the deep learning model independent of the counterfeit features of the data space. In this paper a comprehensive review of data augmentation techniques for object detection is done. Problem of class imbalance is also outlined with possible solutions. In addition to train time augmentation techniques an overview of test time augmentations is also presented.

The design of distributed database has became demanding with the increase in use of IoT and cloud based services. Distributed database system’s performance is totally relies on its design. Allocation of data is one of the major design issues while designing distributed databases. This paper presents a new technique for non-redundant allocation of data in distributed database design. The proposed approach allocates the data based on Simplified Biogeography Based Optimization (Simplified-BBO). The performance comparison of Simplified-BBO based approach is done against the GA and BBO based approaches. The proposed approach helps in decreasing the data communication cost during query execution which results in increasing the overall performance of distributed database systems.

Image segmentation play a vital role in classification. In this paper, to minimize cross entropy, Teacher-learner optimization approach is engaged to explore an optimal sequence of threshold values at distinct levels. One of the most prominent ideas in image segmentation is image thresholding. The suggested scheme is influenced by the shift of advice in the classroom framework, where a leaner hears from the scholar and later corresponds with each alternative.This idea is practiced to search optimal threshold values from fruit images at various levels.The proposed approach exploited the information theory approach called minimum cross entropy. For experiment work, fruit images (red, green, and golden apple) are adopted. PSNR and uniformity performance procedures are exploited to match the evaluation of TLBO - MCET with GA - MCET, and HBMO - MCET.

Image thresholding is one of the most important approaches for image segmentation. Among multilevel thresholding techniques, cross entropy has been widely used by researchers to find multilevel threshold values. In multilevel cross entropy thresholding techniques, main target is to find an optimal combination of threshold values at different levels for minimizing the cross entropy. In this paper, Teaching-Learning-based Optimization (TLBO) algorithm is used to find an optimal combination of threshold values at different levels for minimizing the cross entropy. TLBO algorithm is inspired by passing on knowledge within a classroom environment where students first gain knowledge from a teacher and then through mutual interaction. This new proposed approach is called the TLBO-based minimum cross entropy thresholding (TLBO-based MCET) algorithm. The performance of the proposed algorithm is tested on a number of standard test images. For comparative analysis, the results of TLBO-based MCET algorithm are compared with the results of Firefly-based minimum cross entropy thresholding (FF-based MCET), Honey Bee Mating Optimization-based minimum cross entropy thresholding (HBMO-based MCET) and Quantum Particle Swarm Optimization-based minimum cross entropy thresholding (Quantam PSO-based MCET). Comparative analysis is done based on two most popular objective performance measures, Peak Signal to Noise Ratio (PSNR) and Uniformity. From experimental results, it is observed that the proposed method is an efficient and feasible method to search an optimal combination of threshold values at 2nd, 3rd, 4th and 5th levels. Keywords: Cross entropy, Teacher-Learning-based Optimization (TLBO), Thresholding, PSNR, Uniformity

Object detection is being widely used in many fields, and therefore, the demand for more accurate and fast methods for object detection is also increasing. In this paper, we propose a method for object detection in digital images that is more accurate and faster. The proposed model is based on Single-Stage Multibox Detector (SSD) architecture. This method creates many anchor boxes of various aspect ratios based on the backbone network and multiscale feature network and calculates the classes and balances of the anchor boxes to detect objects at various scales. Instead of the VGG16-based deep transfer learning model in SSD, we have used a more efficient base network, i.e., EfficientNet. Detection of objects of different sizes is still an inspiring task. We have used Multiway Feature Pyramid Network (MFPN) to solve this problem. The input to the base network is given to MFPN, and then, the fused features are given to bounding box prediction and class prediction networks. Softer-NMS is applied instead of NMS in SSD to reduce the number of bounding boxes gently. The proposed method is validated on MSCOCO 2017, PASCAL VOC 2007, and PASCAL VOC 2012 datasets and compared to existing state-of-the-art techniques. Our method shows better detection quality in terms of mean Average Precision (mAP).

In digital image processing, image improvement is an essential step. In this paper, Gaussian, S-shaped and Z-shaped membership functions based on Type-II are used to enhance the image of the fruit. The issue of over and under enhancement can be encountered by fuzzy techniques. Type-II fuzzy has the ability to solve this problem. The suggested approach was applied to photographs of red apple orange and watermelon fruit and found that analytical results were stronger.

Medical images form a vital part of a patient’s record in medical centers. Medical imaging devices generate data with huge memory requirements. Medical image compression is mandatory for storage and communication of medical data for the purpose of diagnosis. Compression removes the extraneous and redundant data in an image to reduce the storage cost as well as data transmission cost. Compression involves removing coding, interpixel or psychovisual redundancy in an image and, at the same time, retaining the integrity of the information required for the diagnosis in medical images. Lossless compression assures exact reconstruction of the original image after decompressing it and provides greater quality but lesser compression ratio. This paper presents two approaches for lossless compression of medical images. In the first approach, Huffman coding is implemented directly on medical images, whereas in the second approach, differential coding is applied on medical images before implementing Huffman coding. Experimental results show that differential Huffman coding improves the compression ratio.

In recent years, analysis and interpretation of video sequences to detect and track objects of interest had become an active research field in computer vision and image processing. Detection and tracking includes extraction of moving object from frames and continuous tracking it thereafter forming persistent object trajectories over time. There are some really smart techniques proposed by researchers for efficient and robust detection or tracking of objects in videos. A comprehensive coverage of such innovative techniques for which solutions have been motivated by theories of soft computing approaches is proposed. The main objective of this research investigation is to study and highlight efforts of researchers who had conducted some brilliant work on soft computing based detection and tracking approaches in video sequence. The study is novel as it traces rise of soft computing methods in field of object detection and tracking in videos which has been neglected over the years. The survey is compilation of studies on neural network, deep learning, fuzzy logic, evolutionary algorithms, hybrid and recent innovative approaches that have been applied to field of detection and tracking. The paper also highlights benchmark datasets available to researchers for experimentation and validation of their own algorithms. Major research challenges in the field of detection and tracking along with some recommendations are also provided. The paper provides number of analyses to guide future directions of research and advocates for more applications of soft computing approaches for object detection and tracking approaches in videos. The paper is targeted at young researchers who will like to see it as platform for introduction to a mature and relatively complex field. The study will be helpful in appropriate use of an existing method for systematically designing a new approach or improving performance of existing approaches.

Background subtraction (BS) is one of most commonly used methods for detection of moving objects in videos that works by subtracting current frame from a background frame. Effective background modeling and threshold plays a crucial role in BS and can govern accuracy and preciseness of object boundaries. This paper proposes a fuzzy entropy based approach modified BS algorithm for moving object detection with Kalman tracker. The standard BS method has been enhanced using concept of fuzzy 2-partition entropy and big bang big crunch optimization (BBBCO). BBBCO has been used to enhance standard BS algorithm for extracting various parameters required in BS algorithm by framing the problem of parameters detection as optimization problem which is solved using concept of fuzzy partition entropy. The proposed algorithm generates optimal threshold values along with various other measures for background modeling. The detected objects are further tracked using Kalman filter based tracker. The evaluation of proposed method has been done on videos from benchmark datasets and statistical parameters have been calculated. The method is also compared with standard BS and another recent study in the field. The results show promise of the proposed method.

Enhancement of hazy images or video is challenging task because of low contrast exhibited in them. Global contrast stretching methods have been successful in restoring contrasts but problems like overcompensation, truncation of pixel values amounting to loss of information tends to creep in. Artifacts may be introduced and images may loose its colorfulness. This paper presents an evolutionary enhancement method for restoring contrast in images or videos while preserving its colorfulness and brightness. The study proposes a novel histogram equalization method inspired by principles of water cycle algorithm. The proposed method first smoothes Y channel of YCbCr color space and divides input frame into two components using Otsu's 2D thresholding. A set of weighing constraints have been formulated and applied to both components individually in a controlled manner. Water cycle algorithm has been employed to exploit an optimal value of weighing factors for enforcement of constraints on individual components. A three dimension objective function has been designed to suitably perform equalization and control enhancement process. Experimental results show that proposed method is effective in removing haze like patterns in images and videos.

Background subtraction (BS) is one of the most commonly used methods for detecting moving objects in videos. In this task, moving objectpixels are extracted by subtracting the current frame from a background frame. The obtained difference is compared against a threshold value to classify pixels as belonging to the foreground or background regions. The threshold plays a crucial role in this categorization and can impact the accuracy and preciseness of the object boundaries obtained by the BS algorithm. This paper proposes an approach for enhancing and optimizing the performance of the standard BS algorithm. This approach uses the concept of fuzzy 2-partition entropy and Big Bang–Big Crunch Optimization (BBBCO). BBBCO is a recently proposed evolutionary optimization approach for providing solutions to problems operating on multiple variables within prescribed constraints. BBBCO enhances the standard BS algorithm by framing the problem of parameter detection for BS as an optimization problem, which is solved using the concept of fuzzy 2-partition entropy. The proposed method is evaluated using videos from benchmark datasets and a number of statistical metrics. The method is also compared with standard BS and another recently proposed method. The results show the promise of the proposed method.

Image segmentation has been used widely for detection and extraction of objects in digital images. Thresholding is one of the effective image segmentation techniques. Fuzzy 2-partition entropy is used effectively for the selection of threshold value. Fuzzy 2-partition entropy required the optimization of some parameters for the selection of threshold value. Teaching-Learning-Based Optimization (TLBO) algorithm has been applied on fuzzy 2-partition entropy to find these parameters and subsequent threshold value. Normally, fuzzy 2-partition Shannon entropy is used for thresholding. In the proposed research work, fuzzy 2-partition Kapur entropy is explored for the evaluation of its potential for selecting optimal threshold value using TLBO algorithm. The present work measures the performance comparison of fuzzy 2-partition Kapur entropy using TLBO with fuzzy 2-partition Shannon entropy using TLBO. The experiments are carried out on standard test images from benchmark dataset. Peak signal to noise ratio, uniformity and structural similarity index are the three different measures which have been used to compare the performance. Results demonstrate that the performance of fuzzy 2-partition Kapur entropy using TLBO is quite promising.

Ductal carcinoma in situ (DCIS) is one type of breast cancer. Clusters of microcalcifications (MCCs) are symptoms of DCIS that are recognized by mammography. Selection of robust features vector is the process of selecting an optimal subset of features from a large number of available features in a given problem domain after the feature extraction and before any classification scheme. Feature selection reduces the feature space that improves the performance of classifier and decreases the computational burden imposed by using many features on classifier. Selection of an optimal subset of features from a large number of available features in a given problem domain is a difficult search problem. For n features, the total numbers of possible subsets of features are 2n. Thus, selection of an optimal subset of features problem belongs to the category of NP-hard problems. In this paper, an attempt is made to find the optimal subset of MCCs features from all possible subsets of features using genetic algorithm (GA), particle swarm optimization (PSO) and biogeography-based optimization (BBO). For simulation, a total of 380 benign and malignant MCCs samples have been selected from mammogram images of DDSM database. A total of 50 features extracted from benign and malignant MCCs samples are used in this study. In these algorithms, fitness function is correct classification rate of classifier. Support vector machine is used as a classifier. From experimental results, it is also observed that the performance of PSO-based and BBO-based algorithms to select an optimal subset of features for classifying MCCs as benign or malignant is better as compared to GA-based algorithm.

Growth of mobile-based applications and traffic on web has made the role of schedulers in communication networks very challenging. The continuous pressure asserted by these applications is so large that even versatile WiMAX networks are facing stiff challenges in fair distribution of resources to these real- and non-real-time applications. Maintenance of fairness and flagrant quality of service levels among various applications is only possible with adaptive and intelligent scheduling mechanism. Design of such system is only possible if artificial intelligence is embedded in base station scheduler. This paper proposes mechanism for allocation of resources in WiMAX networks utilizing fuzzy logic principles. The proposed mechanism guarantees that latency and throughput requirements of real- and non-real-time traffic classes are met. Variations in incoming traffic have been exploited and used in decision making for offering bandwidth to maintain effective performance levels for all traffic classes. The proposed scheduler considers requests from subscribers and exploits powers of fuzzy logic to extract new weights for queues serving various traffic classes in a WiMAX network on the basis of most recent values of latency, throughput and share of traffic. The scheduling framework has been rigorously tested by performing versatile experiments for ensuring quality of service levels under hard hitting conditions. Fairness of scheduler has also been explored and results obtained are quite promising.