IJSRET Volume 5 Issue 2, Mar-Apr-2019



Sensor Based Automatic Fertigation System
Authors:- S. MuthuKumar, M.Naveenraja, V. NaveenKumar, Dr. S.RajaRanganathan

Abstract: – In sensor based automatic system, the irrigation and the fertilizer supplying system can be automated. Our project comprises of two modules irrigation system and the fertilizer supplier system. In irrigation system, the moisture of the soil can be found using the moisture sensor which can be fixed at the different parts of the farm land. For the crops planted, the required water content can be varied with types. Based on the sensor values obtained from the land, the controller will regulate the required amount of water quantity to the crops. This will terminate the motor when the required level reached. In the fertilizer supply system, the fertilizer content present in the soil can be found using the PH sensor fixed at the farm lands. This will find the nutrient content found in the soil. This will supply the fertilizer to the soil based on the sensor values which can be fixed at the farm lands. This will reduce the over watering and over fertilizer supplying to the crops. As the required quantity of water and fertilizer will be supplied to the crops, this will increase the productivity of the crops and improve the texture of the soil.

Predicting Diabetes in Medical Datasets Using Machine Learning Techniques
Authors:- M.Tech. Scholar Arvind Aada, Prof. Sakshi Tiwari

Abstract: – Healthcare industry contains expansive and delicate information and should be taken care of all around cautiously. Diabetes Mellitus is one of the developing amazingly lethal maladies everywhere throughout the world. Medicinal experts need a dependable forecast framework to analyze Diabetes. Distinctive AI methods are helpful for analyzing the information from differing points of view and synopsizing it into profitable data. The openness and accessibility of enormous measures of information will almost certainly give us helpful learning if certain information mining systems are connected on it. The principle objective is to decide new examples and after that to decipher these examples to convey huge and valuable data for the clients. Diabetes adds to coronary illness, kidney malady, nerve harm and visual deficiency. So mining the diabetes information in productive way is a urgent concern. The information mining strategies and strategies will be found to locate the suitable methodologies and systems for proficient grouping of Diabetes dataset and in removing important examples. In this investigation a restorative bioinformatics examinations has been cultivated to anticipate the diabetes. The R programming was utilized as digging device for diagnosing diabetes. The Pima Indian diabetes database was obtained from UCI storehouse utilized for investigation. The dataset was considered and broke down to fabricate compelling model that foresee and analyze the diabetes ailment. In this examination we plan to apply the bootstrapping resembling method to upgrade the exactness and after that applying Naïve Bayes, Decision Trees and k Nearest Neighbors’ (kNN) and think about their execution.

Analysis and Research of System Security Based on Android
Authors:- M. Tech. Scholar Ronak Jain, Prof. Sakshi Tiwari

Abstract: – Android is a savvy versatile terminal working stage center on Linux. Be that as it may, because of its open-source programming and programmable structure character, it drives the Android framework helpless against get infection assaults. This paper has profoundly inquired about from the Linux framework security system, Android-explicit security instruments and other assurance components. Also, on this premise, Android gadgets have accomplished firmly watched on ordinary state. With the goal that aggressors cannot utilize the portion module or center library to get most elevated access consent and be assaulted. In the mean time, to additionally reinforce the security of Android gadgets, it empowers them to appropriately deal with the high-hazard danger. This paper likewise fortified interruption identification framework (HIDS) in light of the host so as to identify noxious programming and reinforce the Android framework level access control.

Analysis of the Impact of Clear Air Atmospheric Conditions on Propagation Loss at Microwave Frequencies
Authors:- Ezeh, Ikechukwu H., Onwuzuruike, V. Kelechi

Abstract: – In the paper, the analysis of the impact of clear –air atmospheric conditions on propagation loss at microwave frequencies is presented. Particularly, the study focused on the impact of variation in point refractivity gradient on multipath fade depth based on the International Telecommunication Union (ITU) recommendation P.530-17 quick planning applications method. The point refractivity gradient depends on the vertical profile of the primary radio climatic parameters, namely, atmo spheric temperature, pressure and relative humidity. Accordingly, the Nigerian Meteorological Agency (NIMET) Radiosonde dataset of the atmospheric temperature, pressure and relative humidity for 2014 in Cross River state were used in the study. The twelve months (2014) Radio sonde dataset of the atmospheric temperature, pressure and relative humidity at the ground level and at height of 65 m above ground were used to compute the refractivity (N) and point refractivity gradient (dN1) for each of the 12 months along with the annual average values of N and dN1. The results show that the ITU annual average value of dN1= –400 is over 200 % of the annual average value of -131.86 obtained from the locally measured data. For each of the 12 months multipath fade depth is computed using the point refractivity gradient obtained in the study along with parameters of a 10 km case study microwave link. The results show that at the frequency of 8 GHz and percentage of time Pw (%) of 0.01 % that a given fade depth is exceeded in the average worst month, the maximum fade depth of 18.95865 dB occurred in October whereas the lowest fade depth of 14.10459 dB occurred in April. Also, the multipath fade depth is inversely proportional to the point refractivity gradient. Furthermore, the result for the multipath fade depth at Pw = 0.01 % and frequencies of 8 GHz , 16 GHz and 28 GHz for the 12 months shows that multipath increase with frequency and the ITU dN1 value continues to maintain the highest multipath fade depth in each of the frequencies considered. Also, the result for the multipath fade depth at the frequency of 16 GHz and Pw of 0.001 %, 0.01 % and 0.1 % for the 12 months shows that multipath decrease with Pw and the ITU dN1 value continues to maintain the highest multipath fade depth in each of the Pw values considered. The ideas presented in this paper is relevant to microwave communication link designers as it enables them to determine the appropriate fade margin that can be specified to accommodate the multipath fade depth that can occur in a given link.

Zigbee Based Routing Security on Home Automation
Authors:- M.Tech.Scholar Amol Ramdas Kasar, Prof. Sakshi Tiwari

Abstract: – Mechanization is a huge field and now daily can be actualized anyplace relying on the application. ZigBee convention IEEE 802.15.4 remote standard can be utilized to set up system for some particular application reason. ZigBee based home robotization remote sensor system can be executed with little starting expense and can be valuable to control fans, lights and other home apparatuses which activity can be controlled with the assistance of a microcontroller. The home computerization can be helpful system for physically incapacitated people just as seniority individual and a similar system can be utilized to control mechanical applications too.

A study on real-time low-quality content detection on Twitter from the users’ perspective
Authors:- M.Tech. Scholar Garvesh Joshi, Prof. Sakshi Tiwari

Abstract: -Location systems of malignant substance, for example, spam and phishing on Online Social Networks (OSN) are regular with little consideration paid to different kinds of low-quality substance which really impacts clients’ substance perusing background most. The point of our work is to identify low-quality substance from the clients’ viewpoint continuously. To characterize low-quality substance conceivably, Expectation Maximization (EM) calculation is first used to coarsely order low-quality tweets into four classifications. In light of this starter think about, an overview is cautiously intended to assemble clients’ feelings on various classifications of low-quality substance. Both immediate and backhanded highlights including recently proposed highlights are recognized to portray a wide range of low-quality substance. We at that point further consolidate word level examination with the distinguished highlights and fabricate a watchword boycott lexicon to improve the recognition execution. We physically mark a broad Twitter dataset of 100,000 tweets and perform low quality content discovery continuously dependent on the described noteworthy highlights and word level investigation. The consequences of our exploration demonstrate that our technique has a high precision of 0.9711 and a decent F1 of 0.8379 dependent on an irregular woodland classifier with continuous execution in the discovery of low-quality substance in tweets. Our work in this manner accomplishes a positive effect in improving client involvement in perusing internet based life content.

Design and Implementation of Internet of Things Based Smart Mirror Using Raspberry Pi
Authors:- Prof.Dhanashree R. Patil, Apurva S. Ghayalkar, Sayali N.Patil, Shweta P. Patil, Pavan P. Patil

Abstract:- In this paper, we describe the design and development of a futuristic mirror that offers simplified and customizable services to the home environment. .It is a small contribution towards smart home. The mirror interface provides users with the versatility needed for better management and integration of daily tasks. On a par with the recent advances in the Internet of Things standards and applications, the mirror is designed to enable residents to control the household smart appliances and access personalized services. The proposed system is set apart from others for its provision of various customized information services for user profile generation.

Fast Image Dehazing Method Based On Homomorphism Filtering and Contrast Enhancement
Authors:- Neeta Joy, Asst. Prof. Jyothi Johnson

Abstract:- Images captured via camera in the bad weather conditions like hazy or foggy ones would result in degradation of images. This is mainly due to the scattering of atmospheric light. This could lead to the deterioration of the image due to the scattering of atmospheric particles and due to the presence of haze. In this paper, a fast algorithm for single image dehazing is proposed using the homomorphism filtering and then using the contrast enhancement technique. Firstly, to accurately measure the atmospheric light an additional channel method is proposed based on quad-tree subdivision. Here the average grays and gradients are used as assessment criteria. Secondly, after finding the atmospheric light portion in the input image filtering is being applied. For this, homomorphism filtering technique is used. This is a generalized technique for both signal and image processing, involving a nonlinear mapping to a different mapping to a different domain in which linear filter techniques are applied, followed by mapping back to the original domain . At last the haze-free image is obtained as the result of atmospheric scattering model. Numbers of both subjective and objective evaluations are being done with the aid of experiments which shows that this algorithm can clearly and naturally recover the image, with at most clarity at the edges where there is sudden changes in the depth of field. Here the algorithmic time complexity is the linear function of the image size. This has advantages in the running time my maintaining balance between the running times and processing.

Commutation Torque Ripple Reduction in BLDC motor Using M-SEPIC Converter
Authors:- K. ThangaPushpam, S. Radhika, I. TamilSelvi, M. VinothKumar, S. Nagaraj

Abstract:- The paper presents new power converter topology to suppress the torque ripple reduction due to the phase current commutation of brushless DC motor. The modified sepic converter is used in a converter circuit it has high static gain for a given duty cycle and step –up-output. Cuk converter has produce the inverted output. The M-SEPIC is a modification of the basic SEPIC. It is superior to the opposite converters each in terms of the input current, purity and potency. It shows very little overshoot or ringing. And also the switching loss is reduced and has low THD value than Cuk converter so it has high static gain and increased efficiency. The theoretical analysis and framework, using MATLAB/ Simulink environment, are given to illustrate the proposed method and the performance is compared with conventional system. And the experimental results are provided for verification of the BLDC drive system.

Design and Implementation of Quad copter
Authors:- Gaurav Gadge, Ankita Ledey, Sourabh Pawer, Gulshan Balpande

Abstract:- The paper presents new power converter topology to suppress the torque ripple reduThis paper presents the design and implementation of an aerial surveillance quad copter for search and rescue application. The aim of this research is to develop a real-time, compact and cost effective drone that will be capable of search and rescue operation and the class is intended to provide the student with both theoretical and practical knowledge in the areas of electronic engineering and design, system integration hardware programming, and control system design and implementation. The frame of the quad copter was made up of very light glass fiber to ensure stability while flying and also to reduce weight of the overall material. The entire design generated a compact and low cost surveillance quad copter with weight of approximately 1.50 kg; which can take photographs from environments with the aid of the onboard mounted camera. Live streaming was done with the help of laptop during fight.

Crop Prediction System
Authors:- Ruchira Nerkar, Bhakti parkhe, Sanjana Bhandure Prof. Bhushan Chaudhary

Abstract:- In recent years, different emerging technologies and applications were introduced for information technology, and now the information technology has become huge part of our day to day life. Especially for agriculture, the yield prediction is very popular among farmers these days. This may contribute to the proper selection of crops for their agricultural development activities based on a farmers experience on a particular field and crop. In earlier stage yield prediction system used data mining techniques in order to predict the crop, analyzed, fertilize and what on. And now-a-days agricultural forecast is becoming more complex and more difficult due to global warming and abnormal weather. To carry on this tasks farmer needs a building of an effective agricultural yield forecasting system on real time. So, crop prediction system is the most fundamental system, which gives soil temperature, soil moisture, humidity, air temperature, of the soil. This helps farmer to analyze which crop should be cultivated, with respect to atmosphere. In this project we use Arduino Uno board to control the entire task which is introduced in programming. And firstly the system is mounted on a particular place in the farm after that the supply is connected to switch board (extension board) and then by switching on the supply it will start and by identifying the parameter’s by the sensors it will displayed on the 16 x 2 LCD screen on system or else by using the router the user can also access the system from anywhere to check the parameters. This system is smart in a technology and will help the future digital world to be more advanced by this system the farmer would get a clear vision about which crop should be cultivated and suitable for the weather.

Leave a Reply

Your email address will not be published. Required fields are marked *