Velammal Engineering College - Department Of Information Technology
V.Chandrasekaran, R.Narendhra Subramaniam, P.T.Hari Narayanan, R.Sri Gokulam
Plant leaves provide sufficient features to distinguish them among other species. Identification of plants using leaf images is a classic problem in digital image processing. In this paper we describe a prosopisjuliflora (informallykaruvelam) identification and elimination system. We developed an Android application that identifies the juliflora plant leaves and activates an injecting machine that contains mera – 71 (glyphosate), 2, 4-D amine and ester which injects the chemical into the plant and clinically kills the plant. The android application is connected with the injecting machine using Internet of Things(IoT).
Venkat Kumari N, Kumar V, JennathPirthouse B, Dr. K. Selva Bhuvaneswari
Agriculture is the backbone of India. By introducing automation into the agriculture sector, the production increases with less production time and minimal labour. In this project, the potatoes are graded based on their attributes such as the size, shape and the external defects. For this purpose, the static image is obtained from the video stream. Several Image processing techniques are applied to collect features such as the Gray level cooccurrence matrix method (GLCM). Multi Attribute Decision Making (MADM) have been adopted for quality evaluation. K-nearest neighbour algorithm (KNN) is used in the classification. The grading accuracy achieved through this scheme is over 90%.
K.Nithya, Ramya.U, Janani.S
Calculating automatic defects detection in MR images is very important in many diagnostic and therapeutic applications. Because of high quantity data in MR images and blurred boundaries, tumor segmentation and classification is very hard. This work has presented one programmed cerebrum tumor location technique to build the precision and yield and diminishing the finding time .The goal is classifying the tissues to three classes of normal, begin and malignant. In MR images, the amount of data is too much for manual interpretation and analysis. Amid recent years, mind tumor division in attractive reverberation imaging (MRI) has turned into a new research region in the field of therapeutic imaging framework. Accurate detection of size and location of brain tumor plays a vital role in the diagnosis of tumor. The diagnosis method consists of four stages, pre-processing of MR images, feature extraction, and classification. After histogram equalization of image, the features are extracted based on Multi-level Complex wavelet transformation (MLCWT). In the last stage, Probabilistic Neural Network (PPN) are employed to classify the Normal and abnormal brain. An efficient algorithm is proposed for tumor detection based on the Spatial Fuzzy C-Means Clustering.
A. Sugashini, R.Swetha,, G. Maheswari
Video streaming is gaining popularity among mobile users. The latest mobile devices, such as smart phones and tablets, are equipped with multiple wireless network interfaces. How to efficiently and cost-effectively utilize multiple links to improve video streaming quality needs investigation. In order to maintain high video streaming quality while reducing the wireless service cost, in this paper, the optimal video streaming process with multiple links is formulated as a Markov Decision Process (MDP). The reward function is designed to consider the quality of service (QoS) requirements for video traffic, such as the startup latency, playback fluency, average playback quality, playback smoothness and wireless service cost. To solve the MDP in real-time, we propose an adaptive, best-action search algorithm to obtain a sub-optimal solution. To evaluate the performance of the proposed adaptation algorithm, we implemented a tested using the Android mobile phone and the Scalable Video Coding (SVC) codec. Experiment results demonstrate the feasibility and effectiveness of the proposed adaptation algorithm for mobile video streaming applications, which outperforms the existing state-of-the-art adaptation algorithms.
T.V.Ashwini, M.Jelinasree, R.Sridharshini, S.Ahamed Ali
In cloud storage benefit, deduplication method is commonly used to decrease the space and transmission capacity prerequisite of administrations by wiping out excess information and putting away just a solitary duplicate of them. Deduplication is the best when different proprietor outsource similar information to the distributed storage, however it raises issues identifying with security and possession. Verification of possession plans permits any proprietor of similar information to demonstrate to the distributed storage server that he claims the information heartily. The various proprietors will transfer the document to the capacity under isolated classification. The verifier will check the document and afterward no one but client can get to the record . The record is produced with mystery key where cryptography system is drawn nearer is utilized. The information clients ask for a similar record which is transferred by the proprietor. To secure the classification of touchy information while supporting deduplication, the cryptography method is utilized to ensure the information privacy and it will produce a mystery key before putting away to the cloud.
A . Rabik, A.Arun Pandi, S.Ajith Kumar, P. Arockia Mary
Identification can be equated to a username and is used to authorize access to a system. Implement Face biometric systems with iterative closest point algorithm and K-NN classification algorithm. And also allow sub-users to access our accounts based on our permission with SMS alerts. Implement Face biometric systems to authorize the person at the time of money transfer or account maintenance. Face identification and resolution technology is crucial to ensure the identity consistency of humans in physical space and cyber space. In current Internet of Things (IoT) and big data situation, the increase of applications based on face identification and resolution raises the demands of computation.
Prema A, Praveen S L, Vamsi Krishna AV, Dinesh Krishnan J
Botnets are the foremost mundane conveyance of cyber-malefactor activity. They’re utilized for spamming, phishing, denial-of-accommodation attacks, brute-force cracking, purloining non-public data, and cyber. A botnet (also referred to as a zombie army) may be a range of network computers that, though their homeowners are unaware of it, are got wind of to forward transmissions (including spam or viruses) to alternative computers on the web. During this paper, we have a tendency to propose a two-stage approach for botnet detection. The primary stage detects and collects network anomalies that are related to the presence of a botnet whereas the second stage identifies the bots by analyzing these anomalies. Our approach exploits the subsequent 2 observations: (1) bot masters or attack targets are easier to find as a result of the impact with several alternative nodes, and (2) the activities of infected machines are a lot of correlative with one another than those of traditional machines.
Devi K, Renuka Devi V, Ramyha V, Pandeeswari.R
Data de-duplication[9] often called intelligent compression or single-instance storage is a process that eliminates redundant copies of data and reduces storage overhead. Data de-duplication techniques ensure that only one unique instance of data is retained on cloud. In this paper, we mainly focus on avoiding duplication in cloud using CP-ABE(cipher text policy -attribute based encryption)and ensure security. CP-ABE[4][8] ensures the high quality randomness in cryptographic key generation with an additional zero-knowledge proof (POK) on the shared cryptographic key and cipher text irrespective of using ABE(Attribute Based Encryption)[1][10]. CP-ABE stores the keys and cipher texts as a pair in the public cloud. Files are listed in the public cloud in order of relevancy to achieve secure de-duplication. File duplication can be identified based on the file content and size. The user can view file information along with key in hidden format. TTP provides key access permission to the requested user, which provides solution for balancing confidentiality and efficiency in performing de-duplication.
S.Nivedhitha, Ruchi V Singh, K.Priyanka
Users find it difficult to fully trust cloud-based services because cloud-based data storage and protection methods are largely user transparent. General encryption schemes protect data confidentiality, but also limit the functionality of the storage system because a few operations are supported over encrypted data. Constructing a secure storage system that supports multiple functions is challenging when the storage system is distributed and has no central authority. In this paper, we propose a shared authority based privacy-preserving authentication protocol (SAPA) to address above privacy issue for cloud storage. In the SAPA, 1) shared access authority is achieved by anonymous access request matching mechanism with security and privacy considerations (e.g., authentication, data anonymity, user privacy, and forward security); 2) attribute based access control is adopted to realize that the user can only access its own data fields; 3) proxy re-encryption is applied by the cloud server to provide data sharing among the multiple users.
Vijayan A, Athithiyan G, Anandaraj K
Due to the rapid growth of computerized techniques in healthcare, it gives raise to the rapid development of computer technology in healthcare for early diagnosis. Especially, recent progress of machine learning in various fields such as marketing, banking, insurance, weather etc. The machine learning will help in predicting the future happenings based on past history. The early diagnosis of any types of disease is not possible at all cases. In order to overcome from this disadvantage, we are introducing the datamining using machine learning algorithms for early detection of disease to prevent loss of life. In this paper, we are discussing about the introduction of machine learning in brain diseases prediction. The decision tree algorithm is used in predicting which type of disease. The types of disease considered are Parkinson’s disease, stress, amnesia, bipolar disorder. The symptoms and physical behavior of the user has been considered. The result has been discussed through the graphical representation.
Dhanvarshinie R, Hamsini R, Varna K Kumar
Network devices are used to congregate data from all over the world, share the assemble over internet for further processing and utilization. This competency serves as the core for Internet of things. IOV has its genesis in the confluence of IOT and mobile internet. But the security issues posed by IOV makes us question its credence. For example, during DDOS, an attacker can deny service to the clients by flooding the network or over consumption of bandwidth. This is where cryptographic verification (private server) and graphical Turing tests(public server) proves useful. They discern legitimate clients and malicious clients. This paper introduces a Query based CAPTCHA technique in VANET with more number of nodes to transmit the packets from client to server. It detects the assailants by using captcha verification during data process. The following parameters are analysed – data delivery rate, efficiency, loss of aggregated data and compares it with the performance of DDOS attack. This project is implemented in TCL, AWK,C++.
Prema A, R.Kavitha Priyadharshini, N S Durga, J Induja
The cloud prosperity is one of the basic characters in cloud, here we would reservation have the capacity to our data into dispersed capacity. A consistently expanding number of clients might need to store their data to PCS (open cloud servers) nearby the quick progression of dispersed registering. Disseminated stockpiling workplaces empower customers to allocate their data to cloud servers to secure neighborhood data amassing costs. Various check errands from different customers can be performed beneficially by the evaluator and the cloud-set away data can be invigorated dynamically. It impacts the clients to check whether their outsourced data is kept set up without downloading the whole data. In our structure we are using the have examining in perspective of the token age. Using this key age methodology take a gander at the key regards from one of a kind key we can find the movements about the record. Novel open affirmations contrive for circulated capacity using as a piece of perceptibility indefinite quality, which requires a lightweight computation on the analyst and delegate most count to the cloud. Not simply set away also the substance will be encoded in the cloud server. If anyone endeavor to hack at the cloud end isn't possible to break the two particular pieces. The security of our arrangement under the most grounded security show. They require first decipher the records and moreover combine the spitted reports from three exceptional regions. This isn't possible by anyone. Anybody can download the intelligences after the waiter with best owner agreement. At the period of download key created (code based key age) and it will send to the archive proprietor. We can download the record need to use the key for check and some extraordinary customers need to download report proprietor approval is crucial.
Vijayan A, V.Kanagalakshmi, G. Pavittra, S.Monisha
In PC organizing, IP address mocking or IP caricaturing is the formation of Internet Protocol (IP) parcels with a manufactured source IP address, with the reason for disguising the personality of the sender or imitating another processing framework. The essential convention for sending information in the Internet arranges and numerous different systems is the Internet Protocol ("IP"). The header of every IP parcel contains the numerical source and goal address of the bundle. The source address is ordinarily the address that the parcel was sent from. By producing the header so it contains an alternate address, an aggressor can influence it to give the idea that the parcel was sent by an alternate machine. With the goal that the IP Spoofing comes into put. Cloud administrations offer better choices for viable sending of an IP trace back framework. We initially introduce novel cloud-based trace back engineering. We have proposed a novel arrangement, named Passive IP Trace back (PIT), to keep away from the difficulties in operation. To catch the sources of IP parodying movement is of extraordinary significance. For whatever length of time that the genuine areas of spoofers are not uncovered, they can't be deflected from propelling further assaults. Indeed, even simply moving toward the spoofers, for instance, deciding the ASes or systems they live in, assailants can be situated in a littler range, and channels can be put nearer to the aggressor before assaulting movement get totaled. The last yet not the slightest, distinguishing the starting points of satirizing movement can help fabricate a notoriety framework for ASes, which would be useful to push the comparing ISPs to check IP source address. The proposed arrangement guarantees that the substance asking for trace back benefit is a genuine beneficiary of the bundles to be traced.Load balancer offers many favourable circumstances over independent Load Balancing and IDS arrangements; it additionally experiences the downsides of both. It has been said that the three things most imperative to a server trough are Security (the capacity to withstand hacking), High-Availability (the capacity to withstand equipment disappointment), and low-idleness (the capacity to benefit asks for rapidly). With the present usage of secure direct, the initial two criteria, security and high accessibility, are fulfilled. Generally concerning idleness, our trials have (indeed) demonstrated that there is an exchange off amongst security and execution. While enhancing the productivity of the programming can relieve this issue to a degree, the exchange off is sure to remain, and picking the correct adjust will probably keep on being a troublesome errand for the framework manager. It is the creators' sentiment that this kind of framework could give a truly necessary extra security device; in any case, a great arrangement of streamlining work stays before it could be broadly conveyed.
Vijayan A, Priyadarshini.S, Rubitha.S
Cloud information proprietors like to outsource archives in a scrambled frame with the end goal of protection saving. Thusly it is basic to create productive and solid ciphertext look methods. One test is that the connection between archives will be regularly disguised during the time spent encryption, which will prompt noteworthy inquiry exactness execution debasement. Likewise, the volume of information in server farms has encountered an emotional development. This will make it much additionally difficult to plan ciphertext seek conspires that can give effective and solid online data recovery on expansive volume of encoded information. In this paper, a progressive grouping strategy is proposed to help more hunt semantics and furthermore to take care of the demand for quick ciphertext seek inside a major information condition. The proposed various levelled approach bunches the records in light of the base importance limit, and after that segments the subsequent groups into subbunches until the point when the limitation on the most extreme size of bunch is come to. In the hunt stage, this approach can achieve a straight computational intricacy against an exponential size increment of report gathering. With a specific end goal to check the genuineness of indexed lists, a structure called least hash sub-tree is outlined in this paper. Investigations have been led utilizing the accumulation set worked from the IEEE Xplore. The outcomes demonstrate that with a sharp increment of reports in the dataset the pursuit time of the proposed technique increments straight though the hunt time of the conventional strategy increments exponentially. Moreover, the proposed technique has leeway over the customary strategy in the rank protection and significance of recovered reports.
Logeswari.M, Varsha.C, Shruthi Kalyani.SP, Gowri.V
The Internet usage is growing everyday along with them there is a significant increase in the security attacks of which the DDoS attack plays a major role. Hackers use different types of DDoS attacks to satisfy their purposes by crashing email servers, websites and other services provided by the Internet. TCP SYN attack which comes under DDoS attack is a great threat in the emerging cyber crime field. This paper focuses on a mechanism to prevent IP spoofing during a TCP SYN attack. The existing system checks the sender for being a blacklisted person or not and executes the prevention mechanism for the transmitted packet accordingly. Though this system prevents the website from the attacks due to IP spoofing a normal user would be blacklisted and would be denied for further service. The proposed system solves the spoofing problem by using a pre-shared key mechanism. In addition this paper proposes a mechanism by which tracing back to the sender node is possible using packet marking technique in certain cases like packet failure or for detecting the sender.
Lavanya.S, Padma.K, Mythili.M, Visu.P
Activity sign discovery and acknowledgment is a vital research subject that constantly keeps more extensive enthusiasm to the exploration in the field of ITS Activity sign recognition (TSR) framework is a noteworthy segment of Intelligent the drivers to drive all the more securely and productively. This paper speaks to another approach for TSR framework where location of movement sign is done utilizing fluffy manage based shading division strategy and acknowledgment is refined utilizing Speeded Up Robust Features(SURF) descriptor, prepared by simulated neural system(ANN) classifier. In the identification step, the locale of intrigue(sign territory) is sectioned utilizing an arrangement of fluffy tenets relying upon the shade and immersion estimations of every pixel in the HSV shading space, present prepared on channel undesirable area. At long last, the acknowledgment of the movement sign is executed utilizing ANN classifier upon the preparation of SURF highlights descriptor. The proposed framework recreated on disconnected street scene pictures caught under various enlightenment conditions. The location calculation demonstrates a high heartiness and the acknowledgment rate is very acceptable. The execution of the ANN show is delineated as far as cross entropy, disarray lattice and recipient working trademark (ROC) bends
Sharon Jennifer E, Jenishiya Hannah. N, Abimathi A, R. Ramya Devi
Cloud computing is an information technology paradigm that enables ubiquitous to share high level services over the internet with minimal efforts. Now-a-days with a rapid usage of cloud most of the organization purchase certain space at cloud server according to their needs.Now, the image features are encrypted and stored in cloud for more secure purposes. After receiving the file, they have to decrypt and find the original file. In our paper, the image file is directly uploaded to the cloud and the features extracted from the uploaded image will be stored in our drive. The automatic key generation is enabled in feature extraction. After uploading, the image seeker can view the broken image from the image selling site. If the image seeker wants to buy the image, he has to send a request to the image owner. The image owner then sends a water mark image to the requested seeker. Image extracting batch verifying algorithm is used required image among the stored image features. After receiving the water mark image, the seeker has to make the payment. The seeker will not be able to download the original image without getting authorization from the owner.
V Subbulakshmi, Ujjwal Anand, Rajesh kannan T, Rajesh N
Twitter slants, an opportune refreshed arrangement of best terms in Twitter, can influence people in general plan of the group and have pulled in much consideration. Lamentably, in the wrong hands, Twitter patterns can likewise be mishandled to deceive individuals. In this paper, we endeavour to explore whether Twitter patterns are secure from the control of malignant clients. We gather in excess of 69 million tweets from 5 million records. Utilizing the gathered tweets, we first direct an information investigation and find confirmation of Twitter incline control. At that point, we learn at the subject level and deduce the key factors that can decide if a theme begins drifting because of its fame, scope, transmission, potential scope, or What we find is that aside from transmission, all of elements above are firmly identified with drifting. At long last, we additionally explore the drifting control from the point of view of bargained and phony records and talk about counter measures. We used an algorithm such as naive Bayesian algorithm is used for machine learning without any explicitly programmed .Behavioural algorithm, finding the behaviour of the person and their emotions. Linguistic algorithm focuses on problem of processing efficiency and ambiguity.
L.Hemalatha, L.Mirakanthan, J.Rajalakshmi
This paper describes hardware developments in a portable air quality sensor assembly for analyzing personal exposure to air pollution. According to the World Health Organization, exposure to air pollution is now the largest single environmental health risk globally. Understanding personal exposure from traditional stationary and spatially disperse monitoring stations is challenging, particularly in urban environments with multiple pollutant sources and complex transport dynamics. Portable air quality sensors have the potential to fill in the gap left by traditional air pollution monitoring. Air pollution sensor technology is decreasing in cost and size, meaning it is now tenable to use low-cost portable air pollution sensors. The data collected by air quality sensor assembly will be stored in cloud using IOT and it can be accessed for future purposes.
M.Ranjini M Arban, R.Vijayarani, V.Mythreyi
Bio metric deals with measurement of human characteristics like face, voice, fingerprint , iris prints, etc . There are devices helps to identify and differentiate person to person based on their bio metrics. These devices consist of transducers which senses and converts the bio metrics into electrical signals and thereby verifying the person’s identity with the data present in the existing databases . This technology is applied in various fields for security purposes , to ensure the entry only to authorized person . As of now it is used in places like military , companies and various institutions for protecting confidential details and to have restricted entry. There are researches that are going on for identifying people even from a distant place , which helps to identify terrorists who live among common people . These sensors are also available in mobile phones which helps to protect the personnel details of the owner . Bio metrics in railway system is already being proposed for implementing in the process of ticket booking few years back but there is no progress for the same. Today there are situations where people find difficulty in booking tickets and there are many cases where people travel without tickets or even ticket black marketing . This paper is deals with a device which uses bio metric system to check if the concerned person has ticket to board the particular train or not.
K.Sundar, S.Faizur Rahuman, R.Depak kumar, M.Vinoth
In cloud computing distributed resources are shared via network in open environment. Hence user can easily access their data from anywhere. At the same time there exist privacy and security issues due to many reasons. First one is dramatic development in network technologies. Another is increased demand for computing resources, which make many organizations to outsource their data storage. So there is a need for secure cloud storage service in public cloud environment where the provider is not a trusted one. In the existing system the owner uploads the file in the cloud using AES encryption it is easily attackable and it is not more secured so if anyone has know the file was encrypted using AES algorithm so they can able to crack that file using some techniques. To overcome this problem we are going to propose a double encryption using AES and RSA algorithm. It was more secured and it is not possible to decrypt a file by someone. Along with this we are adding some features such as Time scheduling and Group sharing .We are using FIFO method in Time scheduling for request and response. Group sharing is when the owner of the file wants to share the file to a particular group of peoples, he can use this technique to share the file to the group he wanted. No other person other than the group members cannot able to access, view or modify the file.
S.Nirmal kumar, R. Dinesh
Manual analysis and decryption of enciphered documents is a tedious and error prone work. Often—even after spending large amounts of time on a particular cipher—no decipherment can be found. Automating the decryption of various types of ciphers makes it possible to sift through the large number of encrypted messages found in libraries and archives, and to focus human effort only on a small but potentially interesting subset of them. In this work, train a classifier that is able to predict which encipherment method has been used to generate a given cipher text. We are able to distinguish 50 different cipher types (specified by the American Cryptogram Association) with an accuracy of 58.7%. This is a 11.5% absolute improvement over the best previously published classifier.Thus this algorithm detects the ciphertext and checks for any threat in the content and blocks the message from being delivered. In addition to that we have modified AES algorithm for time efficiency to encrypt and to decrypt the text at a faster rate.