We all know that there is an exponential growth in the production of digital data. The user of the internet approaches the cloud service providers to upload the video data into the cloud. The cost and size of the digital videos is huge. Hence there is great need for techniques which aims at optimizing the cost and size for storing the data. This paper provides a comprehensive overview of video deduplication system by extracting the global and local features of the video. The major techniques employed in the process of video deduplication are Conv2d i.e., CNN algorithm to extract the local features of the video and hashing algorithms like SHA-256 to extract the global features of the video.
The ability to converse with hearing and deaf persons has always been difficult for those who are tongue-tied. In this paper, we can see different methods which are introduced to help them to communicate effectively. There are many human interpreters or assistant tools to help them communicate, but each person cannot afford that aid. The only mode of communication for them is sign language. Therefore, the project's primary goal is to assist those individuals by providing a system that will recognize the signs, translate them into text, and enable them to lead a normal social life. Previously, a method including hand detection had been developed as a learning tool for novices in sign language. The system was developed using a method based on skin color modeling known as explicit skin-color space thresholding. The specified range of skin tones will distinguish between pixels, or the hand, and non-pixels, or the background. The photos were given as input to a model called the CNN a deep learning algorithm. We will be implementing this project using Keras to train the images. This document provides information on a variety of projects/research on sign language detection in the domains of machine learning, deep learning, and image depth data. This study considers a number of the numerous problems that must be overcome in order to overcome this problem, as well as the future scope.
Underwater images obtained from sea is degraded in its quality due to absorption, scattering, bending of light etc. Red channel compensation is a major process in underwater image restoration which helps in avoiding the effect of red artifacts in the image. In this paper pipelined architecture is developed for red channel compensation unit. The architecture is synthesized for 45nm Technology node and simulated for its functionality verification at 100MHz of operating frequency. The IEEE-754 standard is used to store and access the image from memory. A method is proposed in the designed architecture intended for fast computation and power reduction.
Today as many institutions, corporate organizations, and office campuses lack with an automated Entry control system for unauthorizedvehicles also an efficient automated parkingsystem. This is being done manually with security people, who must go to each vehicle at the main gate and verify, whether the person in the vehicle belongs to that organization or not. Which is really a tediousjob, wastes time, causing congestion at the gate making the authorized person belonging to the organization toexperience the annoying delay at the main gate and also in parking their vehicle. The proposed RFID based IoT solution system in this paper tries to resolve this by allowing the security at the main gate for easily recognizing the authorized employee vehicle and making employee to get the message with appropriate parking slot number to his phone at the entrance itself. This completely avoids the traffic congestion at the gate and the problem of searching for parking slot in the campus or basement. This system also provisions the security staff to have the information on authorized and unauthorized vehicles with the vehicle number and entry time, and manage the parking space or slots effectively.
Hardware inventory management plays a pivotal role in warehouse goods management for several industrial activities. Hardware inventory management helps to track warehouse goods, ensures that there is an adequate amount of goods left back in the inventory, and prevents overstock or shortage of stocks. This paper discusses the above procedure to be implemented with the help of IoT. In this study, the purpose of the application was to serve withless wiring, fewer components, and to make the system be available at a low cost. This system uses NodeMCU ESP8266 module which is an open-source platform having built in Wi-Fi-module feature, which is best suited for IoT related applications. The tracking and stock details were automatically updated on a management systemconnected to the same network as that of the developed system. A detailed graphical representation of the stock availability in a warehouse can be viewed. From the graph, the inventory manager can decide whether to increase the stock quantity or not. The methods that are currently being followed are more human- dependent and not framed with modern technology. Therefore, inventory details are not always accurate, which may also be one of the reasons why a business experiences loss. With this system, the entire process is made more automated, with less human intervention, thereby providing accurate values for inventory goods.
Flooding is more than a momentary influx of water onto ordinarily dry terrain. Floods are the most frequent natural calamities in certain states, like Assam, Kerala, Tamil Nādu, and Bangladesh. As per the IPCC-2022 third report, mostly 44 percent of environmental disasters globally are reported yearly, flash floods represent 22 percent of all economic damages globally, and the severity of death rate is about 10 percent worldwide. Failing to evacuate flooded regions or coming into flood waters can result in harm or death. By utilizing the different correlated data, such as the availability of water from all resources, including canals, rivers, glaciers, precipitation in that area, and past rainfall data the prediction will be accurate. This survey is related to flood forecasting (FF) based on precipitation and water obtainability index (WOI) using machine learning (ML) and deep learning. Deep Learning (DL) has become an evolutionary and adaptable technique that revolutionizes business applications and produces new and improved model creation and scientific discovery capabilities. although dl adoption in hydrology has so far been sluggish, the time is now right for innovations.
When the globe was hit by the vicious Covid 19 pandemic, multiple industries faced the virus's wrath and that included the agricultural warehouse industry. Consequently, many warehouses which had received large shipment stocks of agricultural products were never to be used again as it had reached its expiration date. This led to major losses for the agricultural warehouses as well as losses in crops for farmers and large scale agriculturists. The main objective of this paper is to build a model which utilises 3 heavy-weight algorithms (Seasonal Autoregressive Integrated Moving Average - SARIMA, Long short term memory - LSTM and Holt Winters) and predicts the agricultural needs of retailers and consumers based on previous data from different warehouses. Deploying this system will not help in the regulation of goods in warehouses but will also aid in maximizing the profits and minimizing the losses for warehouses. The algorithm with the least MAE(Mean Absolute Error) value will be considered for forecasting the sales of the aforementioned product.
Recently, it has become simple to produce trustworthy face video exchanges that leave a few signs of deception thanks to in-depth free reading software tools (DF). Despite decades of effective use of visual effects in digital video deception, recent developments in in-depth learning have significantly improved the genuine nature of misleading content and the accessibility that can be achieved with it. This is referred to as AI-synthesized media or DF in short. Making DF is a simple task that uses practical tools. However, it is a significant difficulty if these DFs are discovered, because it is hard to train the algorithm for identifying DF. CNNs and RNNs have helped us come closer to DF. The Convolutional Neural Network (CNN) is used by the system to extract features at the individual level. The continuous neural network (RNN) states learn to recognize whether or not a video is being deceived and be able to spot temporary anomalies among the frames given by DF's creative tools thanks to these capabilities. An extensive collection of pseudo-videos gathered from a common data source is the anticipated outcome. We demonstrate how our method can produce a competitive outcome in this work that is simple to utilize.
Road accidents in the country continue to be a leading cause of death, disabilities and hospitalization despite our commitment and efforts. Road accidents happen at unexpected moments and thus cannot be anticipated. Lives are lost as a result of the inability to provide medical assistance on time. A small delay in notifying the concerned authorities can result in delayed medical help, which in turn results in loss of life. Hence, there must be a proper system that notifies nearby health centers as soon as an accident is detected. We aim to develop such a system which has the potential to so many lives. The solution which is proposed in this paper involves collision sensors to detect when an accident occurs, and then immediately alerting the emergency center. When an automobile meets with an accident, the microcontroller detects the collision with the help of sensors and immediately sends a voice message to the emergency center. This message contains GPS coordinates and timestamp of the accident. A camera module, which captures images periodically immediately after the accident, will be placed in the automobile. These images captured are then sent to the emergency center. These can be used to assess the severity of the accident and decide the best course of action.
One of the most crucial problems with artificial intelligence systems is thought to be the identification of correct kidney diseases through machine learning. Manual diagnosis for predicting the kidney disease by doctors is time consuming and may raise the workload on doctors. So, the developed system uses a machine learning technique for predicting the chronic kidney disease which may help the doctors in early prediction of the kidney disease. In order to diagnose chronic kidney disease four Machine Learning technique namely Naïve Bayes, Random Forest, Decision Tree and Support Vector Machine is used. Naive Bayes uses probability to forecast kidney disease, whereas decision trees are used to generate categorized reports for the disease. This system will compare the accuracy score of each Machine Learning technique. Hence, Random Forest gives the better performance compared to other classification methods with accuracy score of 98.75 % , $\mathbf{F1}=\mathbf{score}=\boldsymbol{99\%},\mathbf{ Precision}=\boldsymbol{99\%},\mathbf{Recall} =\boldsymbol{99\%}$ . This paper shows the efficiency and accuracy of the predicted chronic kidney disease.
In communication networks, secure connectivity is critical. The failure of a dominant node may jeopardize the network's stability. As the network grows in size, it becomes increasingly difficult to maintain track of all of its dominant nodes and secure connections. As a result, large network analysis can be performed quickly and easily using a computer system that understands networks as matrices, and a fit algorithm will make calculating work simple and quick. In this paper, we present two algorithms for determining the dominance number and secure dominance number of a fuzzy network, with the objective of comparing the concepts of dominance and secure dominance in a fuzzy structure to diagnose the patient symptoms which guide them to appropriate medical treatment.
Human blood contains RBCs, WBCs, platelets, and plasma. The counting of blood cells is important since it shows one's overall health. Counting Blood Cells encompasses all of the cells required to evaluate a person's health and diagnose illnesses. A lack of red blood cells (RBC), which account for 99 percent of all blood cells, causes a number of blood disorders. The typical wait time for blood results and reports in today's hospitals and clinical laboratories is currently 24 hours in cases of high-severity diseases with high mortality rates. For high-risk diseases such as hepatitis B, doctors and healthcare technicians advise patients to wait as little as possible and to begin treatment as soon as possible. The process starts with picture acquisition and image enhancement, which reduces noise in photos while preserving the edges, before converting them to binary images and separating the region of interest from the background. RBC is separated from the other blood components during the segmentation process. The blood image is morphologically treated before RBC counting is performed using the Hough transform, a powerful image segmentation technique.
In a fast technology growing world with respect to the data, information security is the major criteria. To have the robust system the security is the foremost task. The personal identification system and biometrics are playing the important part in security of the system. To secure the system, the biometrics, OTP generation, irisdetection etc., are some of the securities concerned aspects are taken. Biometrics security system is one of the old security systems. During the biometric security systems, the main challenge is acquiring images with Visual Wavelength or Near-Infrared lighting in limited and unconstrained situations. This paper proposes a unique approach for personal authentication based on the merging of periocular and iris sensors. This approach may be used in the ITS (Intelligent Transport System). In the ITS, depending on the identification of the passengers, only privileged or the authorized passengers can be transported in the vehicle. The suggested system includes components for feature learning for categorization after image pre-processing. During image pre-processing, a slicedannular iris from an ocular image is turned into a remediedimage region. The local periocular region was extracted using iris localization parameters. Images suffer from varied noise abnormalities. Depending on the abnormality of the image, various data inadequacy and complicating situation arises. To deal with this problem, a novel data augmentation technique has been developed.
Deep learning (DL) techniques like recurrent neural networks (RNN) and convolutional neural networks (CNN) are currently being utilised to improve management tooling and workflow classification to increase operational effectiveness. Reliability could be increased, but because of CNN's intricacy, actual research is therefore limited. A brand-new DL structure is suggested in this study to incorporate the visualization of mappings (IVM) within Masked R-CNN. During the first approach, this paradigm, IVM-CNN combines the best features of both approaches, including (1) IVM for object tracking by emphasizing geospatial data for sector recommendations and (2) CNN for machine vision by relying on data for picture categorization. Using spatial and temporal statistics along with visual functionalities, the said approach is tested on M2CAI 2016 contest sets of data, outperforming all prior creations and accomplishing futuristic outcomes to 97.1 mAP for device diagnosis and 96.9 mean rate. It also performs at 50 FPS, which is ten times quicker than region-based CNN. Masked R-CNN substitutes the region proposal network (RPN) with a region proposal module (RPM), which more precisely generates boundary boxes and reduces the demand for labeling. Microsoft HoloLens software is also being generated to offer an augmented reality (AR) stationed approach for clinical education and help.
With the ever-increasing number of diseases in today's world, there is a need for a system to provide early diagnosis and the root cause of human health. Indian and Chinese traditional medicine system provides natural and simple solutions to detecting health issues. Nadi-Nidan is an ancient medical technique, traced back to ancient Indian traditional health monitoring, known to indicate all the health features of a human body. In Nadi Nidan, Wrist pulses or arterial pulses are sensed to diagnose the health status. The study was carried out to design a non-invasive system for wrist pulse analysis that gives us the heartbeat, IBI (Inter-Beat-Interference) and the body type, to support doctors in routine diagnostic procedures and provide detailed procedure for obtaining the complete set of the Nadi signals as a time series. An Ayurveda practitioners and physicians can use this prototype for pulse reading and uniformate in analysis. The proposed model specifically deals with data acquisition of three Nadi signals Vata, Pitta and Kapha. Signals are obtained by using PPG sensors. Arduino is used as the data acquisition hardware. Identification of Prakruthi of the subject was carried out based on the amplitude of Vata, Pitta and Kapha signal acquired at the wrist and achieved 83% accurarcy. Vata, Pitta and Kapha of diabetic and normal subject were analyzed.
Lung disorders can be fatal if not treated in the right manner. Symptoms of respiratory disorders include wheezing, breathlessness or difficulty in breathing, cough, hoarseness, and chest pain to name a few. Although the symptoms look common, lung disorders often go undetected due to various reasons such as misdiagnosis, expensive diagnostic techniques, lack of awareness and negligence. In most cases, the patient is required to take a pulmonary function test which includes thoracoscopy, chest imaging (X-rays), electrocardiography and bronchoscopy. In this paper we explore an alternative technique for detecting respiratory disorders through analysis of lung sounds. Lung sounds are vital factors of respiratory health and disorders. They are produced due to the movement of air and secretions in lung tissue or they might also be generated due to the presence of any infection or anomalies. Asthma or Chronic Obstructive Pulmonary Disease (COPD) patients often wheeze as a result of an obstructive airway disease. These sounds can be captured using digital stethoscopes which can then be converted into audio signals for further processing. This audio data gives us the opportunity to diagnose respiratory disorders like pneumonia, asthma and bronchiolitis using deep learning techniques such as convolutional neural networks. We also propose a design for the digital stethoscope which can help record lung audio samples.
In modern times, Security and surveillance in households have become somewhat of an especially important necessity. Advancement in technology has made it possible for everyone to access or break into different houses easily. The main purpose of our project is to build an advanced surveillance system that can be used to detect the different faces or any movement that may occur while in the view of the surveillance camera. This system is also supported by an application that has unique features to make it more user friendly for the users. Not only is the user notified when an unauthorized entity is detected, the user is also allowed to add different faces or objects that will be ignored during the process of theft detection. This has been achieved using unsupervised machine learning where a given set of data is compared with the actual live feed from the surveillance camera to check for any anomalies in its surroundings. The Dataset or the data used in the proposed system are a few images in the format of. JPEG and. JPG which can be stored in the given location manually by the user or through the application itself. The proposed model recognizes the images in any of the available formats. The modules used in this system are powered by a strong python module named Open CV. This module supports various face recognition algorithms such as Haar Cascade, Eigen Faces, Fischer Faces, Local Binary Pattern Histogram (LBPH), etc and this module is responsible for all the image recognition, classification, and identification. The images extracted from the dataset are real time Image frames obtained from the user webcam, both are compared using the Face Recognition module in python which uses the Regions for - Convolutional Neural Network Algorithm (R-CNN) and Unsupervised learning approach to detect and differentiate between objects in Real Time. This system also includes a message transmitting feature which works with the help of the Simple Message Transfer Protocol (SMTP) module in python. Whenever an unknown user is identified by the system an email is sent to the admin or the user using the SMTP message transfer module which registers the mail address of the user when the initial setup of the system takes place. Hence a robust, secure and user-friendly device is developed that can always keep your house theft free.
Water degradation has become a critical theme of concern in recent years. Water is necessary for biological species survival and living activity is strikingly dependent on the quality of the water (i.e., physical, chemical, and biological aspects of water). The aim of the paper is to develop Internet-based Water Quality Monitoring System to determine the water quality parameters namely turbidity, PH, temperature etc. The developed model encompasses ESP32 Wi-Fi & Bluetooth Microcontroller with appropriate sensors and communication circuitry. The paper proposes a cost effective Remote Operated Underwater Vehicle which can monitor the parameters successively for prolonged period. The developed model is tested for three different cases and the parameters inferred are communicated through Thing Speak analytics platform
Mental State of the subject is evaluated by using EEG signals. EEG signal from the brain is taken by using Mind wave kit, which gives the raw EEG waves by the non-invasive method only using single electrode. To induce emotion in to the subject different emotional videos are shown and respective emotion EEGs are collected. so the electrical different EEG wave Alpha, Beta, Delta, Gama and theta varies, different waves having its nativity according to the emotional changes. Lucid scribe toolkit support to collection of data from the Mobile mind wave, the data exported to the excel and by finding the minimum and maximum value of every EEG wave, this is in the numerical values with known Mental states are set with numerical values like 0 neutral, 1 Happy, 2 Disgust, 3 Sad, 4 Angry. In this work Mental state evaluation using signal processing is carried by preparing 280 datasets are prepared by showing them different videos related to respective emotions. By using neuro sky's Mindwave kit brain waves are recorded at the forehead values are tabulated accordingly. 280 datasets are fed into Orange, open-source machine learning and data visualization module and algorithms are compared by extracting confusion matrices.
Augmented reality is a live direct or indirect view of physical-word. ‘Augment’ by computer generated or extracted real word sensory. Augmented Reality supplement your word with digital object of any sort. It combines physical and virtual word. This is one of the immersive technologies that will change your lifestyle in the near future. AR technology has the positive opportunity and service to educate change. Agriculture has been an important source of food and has always been a very important aspect. Agriculture is highly labor-intensive and highly dependent on the knowledge of individual farmers, causing management problems. It can make a decisive input to the best possible operation management of AR. Improve farmers reality in insect research and pest control AR correction compared to orthodox techniques and teaching methods (circles, speeches, etc.). In the eyes of the overall public, agriculture has a tendency to be easier. This is just like the easy case of sowing and reaping seeds. Agriculture is absolutely the becoming a member of collectively of diverse sciences and is a completely complicated manufacturing system. The demanding situations dealing with in enterprise appear like exacerbated with the aid of using the extra of a well-skilled team of workers in at the upward thrust countries. It isn't unusual for uneducated humans to interact in agriculture the use of suspicious archaic techniques. Not surprisingly, they fail, which ends up in diverse demographic demanding situations for society.
