h2 { margin-top: 0.42cm; margin-bottom: 0.42cm; color: rgb(0, 0, 0); page-break-after: auto; }h2.western { font-family: “Liberation Serif”, serif; font-size: 12pt; }h2.
cjk { font-size: 10pt; }h2.ctl { font-family: “FreeSans”; font-size: 12pt; font-weight: normal; }h1 { margin-bottom: 0.42cm; color: rgb(0, 0, 0); text-align: center; page-break-after: auto; }h1.western { font-family: “Liberation Serif”, serif; font-size: 14pt; }h1.cjk { font-size: 14pt; }h1.
ctl { font-size: 14pt; }p { margin-bottom: 0.25cm; line-height: 120%; }ABSTRACTApplication for detecting an object using themobile video camera and giving voice instructions about the currentlocation of the blind user by using GPS and to give the direction ofan object to the blind person. User will need to train the systemfirst regarding the object information. System extracts the featuresto search objects in the camera view to know the direction of object,where it is placed using angle extraction feature. This Androidapplication gives warning of the obstacles in the way to the user. .
Also the proposed system converts the text intoaudio for giving the instructions about the directions to the blindperson and for such conversion the Speech synthesizer technique getsused. The camera of the phone is enough for this purpose and nospecial hardware is required, ensuring that it requires minimaleffort from the user to use the application during everyday life.System gets used in social approach where the object in place or inpath everyday life and with the help of this system blind personeasily travel or visit common places such as school, college,hospital, shopping mall and travel on roads. INTRODUCTION Object recognition Blindpeople face several problems in their life, one of these problemsthat is the most important one is detection the obstacles when theyare walking. In this research, we suggested a system with two camerasplaced on blind person’s glasses that their duty is taking imagesfrom different sides. By comparing these two images, we will be ableto find the obstacles. In this method, first we investigate theprobability of existence an object by use of special points that thenwe will call them “Equivalent points”, then we utilizebinary method, standardize and normalized cross-correlation forverifying this probability.
Best services for writing your paper according to Trustpilot
This system was tested under threedifferent conditions and the estimated error is acceptable range. Optical character recognition Machinereplication of human functions, like reading, is an ancient dream.However, over the last five decades, machine reading has grown from adream to reality. Optical character recognition has become one of themost successful applications of technology in the field of patternrecognition and artificial intelligence.
Many commercial systems forperforming OCR exist for a variety of applications, although themachines are still not able to compete with human readingcapabilities. 2 LITERATURE REVIEW In 2012, Nidhi describes Prototype system for color based objectdetection is successfully implemented and tested. The test resultsshow that the detection method used in the paper can accuratelydetect and trace any object in real time.
This paper shows themethods of Image processing and detecting an object in it based onits specific color, by using Open cctv real time implementation ispossible. Thresholding of the generated image is necessary in orderto segment the image pixels and let them free from each other.In 2012, Sanjivani Shantaiya presents an extensive survey of objectdetection approaches and also gives a brief review of each approach.Various object detection approaches are discussed as feature based,template based, classifier based, motion based as per the reviewedpapers.In 2013, Shalin A.
Chopra tells about OCR system for offlinehandwritten character recognition. The systems have the ability toyield excellent results. Preprocessing techniques used in documentimages as an initial step in character recognition systems werepresented. The feature extraction step of optical characterrecognition is the most important.
In 2013, Khushboo Khurana describes In this paper, we have discussedvarious object detection techniques. The template matching techniquerequires large database of image templates for correct objectrecognition. Hence it must be used only when limited objects are tobe detected.
Global features and shape based method can give betterresult and are efficient as compared to local features. In 2014, Sukanya C.M describes In this survey paper all the mainterminology of object detection have been addressed. These includeobject detection methods, feature selection and objectclassification. Most commonly used and well recognized methods forthese phases have been explained in details.
Different methods forobject detection are like frame difference, optical flow andbackground subtraction. In 2015, Chirag Patel describes Although Tesseract is command-basedtool but as it is open source and it is available in the form ofDynamic Link Library, it can be easily made available in graphicsmode. The results obtained in above sections are obtained byextracting vehicle number from vehicle number plate. In 2015, Sean O’Brien After successfully creating our Hungariancharacter and language models, we assessed the accuracy of theOCRopus software. We compared the results of our models versus thedefault English models on a Hungarian algebra textbook written in1977 by László Fuchs. In 2016, Mayuri B Gosavi In this system, we have proposed anartificial neural network-based simple colour and size invariantcharacter recognition system to recognize English alphanumericcharacters. Our proposed system gives excellent result for thecharacter letters when they are trained and tested separately butproduce satisfactory result when they are processed together. In 2016, Astha Gautam With the help of object recognition concept wecan simply identify the objects present in an image or a videosequence.
There are number of techniques and methods that can beapplied for having the desired result. In 2016, Sukhpreet Singh This paper has presented a related work onEnglish OCR techniques. Various available techniques are studied tofind a best technique. But is found that the techniques which providebetter results are slow in nature while fast techniques mostlyprovide inefficient results. 3 PROPOSED SYSTEM The main objective of this project is todevelop an application for blind people to detect the objects invarious directions, detecting pits and manholes on the ground to makefree to walk Detecting objects using image processing can be used inmultiple industrial as well as social application. This project isproposing to use object detection for blind people and give themaudio/ vocal information about it.
We are detecting an object usingthe mobile camera and giving voice instructions about the directionof an object. User must have to train the system first about theobject information. We are then doing feature extraction to searchfor objects in the camera view.
We are taking help of angle whereobject is placed to give direction about the object.Figure3.1:Systemarchitecture of ODR 4 CONCLUSIONOptical character recognition technologytoday can read and recognize an array of languages and convert filesinto a number of formats. Even though it was developed decades ago,it continues to be changed, edited and improved. The OCR applicationis an apt tool for picture to text conversion and can be used bydifferent people in a variety of environments. It continues to besupported by a wide range of products and systems, fromtop-of-the-line machinery to compact easy-to-use solutions.All throughthe years, the methods of OCR systems have improved from primitiveschemes suitable only for reading stylized printed numerals to morecomplex and sophisticated techniques for the recognition of a greatvariety of typeset fonts and also hand printed characters.REFERENCES1.
Arica, N., Vural, F. T.Y.
, An Overview of Character Recognition focused on OfflineHandwriting, IEEE Transactions on Systems, Man and Cybernetics –Part C: Applications and Reviews, 31(2), pp 216–233, 2001.2. Bunke, H., Wang, P.
S. P. (Editors), Handbook of CharacterRecognition and Document Image Analysis, World Scientific, 1997. 3. Chaudhuri, A.
, Some Experiments on Optical Character RecognitionSystems anguages using Soft Computing Techniques, Technical Report,Birla Institute of Technology Mesra, Patna Campus, India, 2010.4. Cheriet, M., Kharma, N., Liu, C. L., Suen, C. Y.
, CharacterRecognition Systems: A Guide for Students and Practitioners, JohnWiley and Sons, 2007. 5. Dsholakia, K., A Survey on Handwritten Character RecognitionTechniques for various Indian Languages, International Journal ofComputer Applications, 115(1), pp 17–21, 2015. 6. Mantas, J.
, An Overview of Character Recognition Methodologies,Pattern Recognition, 19(6), pp 425–430, 1986. 7. Rice, S. V., Nagy, G., Nartker, T. A.
, Optical CharacterRecognition: An Illustrated Guide to the Frontier, The SpringerInternational Series in Engineering and Computer Science, SpringerUS, 1999.
