Intelligent Voice Stick for Blind and Partially Sighted People
Moumita Das, Sahadev Roy, Pinaki Chakraborty, C.T.Bhunia
Department of ECE, NIT, Arunachal Pradesh
Yupia, India
Abstract- The smart electronic voice stick is intended to provide overall measures- artificial vision and object detection. This application is to convert the text written in language to speech. This application helps user to type text and as output it gets the speech or it may be scanned paper that is converted to speech and the user gets its speech output. The aim of the overall system is to provide a low cost and efficient navigation aid for blind which gives a sense of artificial vision by providing information about the environmental scenario of static and dynamic objects around them.
Ultrasonic sensors are used to calculate distance of the obstacles around the blind person to guide the user towards the available path. Output is in the form of voice which the blind person can hear e.g., right, left etc.
Device is a portable text scanning tool, utilizing the OCR function to identify text and convert this information into voice. This advanced technology allows visually impaired people to read everything from books and newspapers and study materials and they can walk and move everywhere from one place to another place without help of other persons.
Keywords- Android, OCR, Algorithm, Voice Stick, Ultrasonic sensor.
I. Introduction
Vision is the most important part of human physiology as 83% of information human being gets from the environment is via sight. The 2011 statistics by the World Health Organization (WHO) estimates that there are 285 billion people in world with visual impairment, 39 billion of which are blind and 246 with low vision. The traditional and oldest mobility aids for persons with visual impairments are the walking cane (also called white cane or stick) [1]. The most important drawbacks of these aids are necessary skills and training phase, range of motion and very little information conveyed. With the rapid advances of modern technology, both in hardware and software front have brought potential to provide intelligent navigation capabilities [2]. Recently there has been a lot of Electronic Travel Aids (ETA) designed and devised to help the blind navigate independently and safely. Many blind guidance systems use ultrasound because of its immunity to the environmental noise.
Another reason why ultrasonic is popular is that the technology is relatively inexpensive and also ultrasound emitters and detectors are small enough to be carried without the need for complex circuit.
II. Aim of the project
The main objective of this project is to provide artificial guidance to the visually impaired people with the help of Ultrasonic Sensors, a speaker physically mounted on a stick. The blind persons can detect the obstacle in front of him using the sensor and the sensor will detect how far it is (obstacle) from the blind person. The blind persons also are able to read books, newspapers, and study materials using this application. There will be a scanner that will scan the whole document and from the scan copy the blind persons will be able to listening the document through the speaker as output.
III. SIGNIFICANCE OF THIS WORK
The significance of this project is to help the visually impaired people with appropriate voice commands that are played through the speaker. The scanner will scan the whole document form the materials and these scanned documents will be announced through the speaker that will be listening by the blind persons and the ultrasonic sensor will detect the obstacle in front of the blind person from a certain distance and speaker will declare how far it is and in which side it is situated of the blind person or partially sighted people.
IV. Related work
Over the last decades, research has been conducted for new devices to design a good and reliable system for visually impaired persons to detect obstacles and warn them at danger places. There are some systems which has some deficiencies. An obstacle avoidance wearable portable computer which is only for indoor navigation [3]. It was equipped with two modes, in the first one the system information was translated to audio in different sounds. One sound for free for travel direction and other for blocked, it was difficult for the person to differentiate the sounds. Other problem was the system would not know the user momentary position. S. Innet and N. Ritnoom have introduced a stick for distance measurement using ultrasonic sensors, which is a complex and time wasting process. The stick has different vibration modes for different range which is difficult for a blind to differentiate, it needs time for training. The stick informs the person clearly at dangerous stage which conveys less information and safety [4]. The stick has no location and positioning features. J.Na proposed an interactive guide system for indoor positioning, which can’t detect the obstacles and hurdles. The system is not suitable for the outdoor activities.
V. Problem Statement
Visually impaired persons have difficulty to interact and feel their environment. They have little contact with surrounding. Physical movement is a challenge for visually impaired persons, because it can become tricky to distinguish where he is, and how to get where he wants to go from one place to another. To navigate unknown places he will bring a sighted family member or his friend for support. Over half of the legally blind people in the world are unemployed. Because limited on the types of jobs they can do. They have a less percentage of employment. They are relying on their families for mobility and financial support. Their mobility opposes them from interacting with people and social activities. In the past different systems are designed with limitations without a solid understanding of the non-visual perception. Some of the systems are only for indoor navigations, and has no hurdle detection and determining location feature in outdoor environment. There is no one system available to navigate indoor, outdoor and also determine location and position to easily facilitate the visually impaired persons. The available systems are very costly; some of the systems are very heavy cause physical fatigue and required training to use.
VI. Literature survey
a. Optical Character Recognition
Optical character recognition is usually abbreviated toOCR is themechanicalorelectronicconversion of scanned or photographedimagesof typewritten or printed text into machine-encoded or computer-readable text. It is widely used as a form of data entry from some sort of original paper data source, whether passport documents, invoices, bank statement, receipts, business card, mail, or any number of printed records. It is a common method of digitizing printed texts so that they can be electronically edited, searched, stored more compactly, displayed on-line, and used in machine processes such asmachine translation,text-to-speech, key data extraction andtext mining. OCR is a field of research in pattern recognition,artificial intelligenceandcomputer vision.
Early versions needed to be programmed with images of each character, and worked on one font at a time. "Intelligent" systems with a high degree of recognition accuracy for most fonts are now common. Some commercial systems are capable of reproducing formatted output that closely approximates the original scanned page including images, columns and other non-textual components.
It could recognize text printed in virtually any font that the best application of this technology would be to create a reading machine for the blind, which would allow blind people to have a computer read text to them out loud. This device required the invention of two enabling technologies– theCCDflatbed scannerand the text-to-speech synthesizer. In the 2000s, OCR has been made available online as a service (WebOCR), in acloud computingenvironment, and in mobile applications like real-time translation of foreign-language signs on aSmartphone.
Various commercial and open source OCR systemsare available for most commonwriting systems, including Latin, Cyrillic, Arabic, Hebrew, Indic, Bengali (Bangla), and Devanagari, Tamil, Chinese, Japanese, and Korean characters.
Character recognition system is classified into two recognition system: online recognition system and offline recognition system. The online recognition system is found to be superior to the former as the initial step of determining the location of text is not required. In offline character recognition the scanned document is taken and processed by means of recognition algorithm. It is a challenging task due to variation in shape, font size, and document quality of the character.
b. Obstacle detection unit
Electronic voice sticks have been classified in three classes:
· Obstacle detectors
· Clear-Path indicators and
· Environmental sensors
The first class is based on sensory or artificial vision systems. The sensory systems emit ultrasonic to its path which is reflected by the obstacle; the system calculates the distance from the object according to the time difference between the emitted and received sound.
The proposed system uses an array of ultrasonic sensors which basically works on the principle of the ultrasonic sound generation and alert mechanism [6], [7]. The system is however having a dual feedback mechanism i.e. it has an additional vibratory feedback mechanism. This enhances the overall feedback received by the blind user who receives the outputs generated in different formats of vibration i.e. high, low, medium and strong vibrations.
VII. methodologies
The sensor based circuitry consisting of sensors Ultrasonic sensor is used to detect ranges from obstacles. Ultrasonic Obstacle sensor is applied to the voice stick. The microcontroller reads these sensors and drives a buzzer, a LED. An audio output is designed by a buzzer alarm.
The output indications provided by the microcontroller are distinctive as per sensor. Based on the strength of the beeping of the buzzer or the blinking of the LED embedded with the stick a disabled person may determine if they are walking towards a manhole or an edge or a large opening at nearby bottom or something similar. At the same time they may get the sense of their distance from nearby objects and if they can walk in a wet or muddy or potentially slippery terrain.
VIII. system design
The proposed design for smart voice stick distinctly consists of three units:
a) The GPS Unit
b) The Obstacle Detection Unit
c) GSM Unit
Figure: 1
System Design
IX. Algorithms
Algorithm 1:-
Step 1:-Start
Step 2:-Scan the handwritten document
Step 3:-Take one character at a time
Step 4:-Extract features of the character
Step 5:-Print the character
Step 6: Start traversing the sentence given in Bengali.
Step 7: Match the first word with the words in the roots lexicon.
Step 8: Translate the Bengali word into corresponding English word from the
Words stored in the dictionary.
Step 9: Find out the attributes of the word found by analysing the 16 bit tag vector.
Step 10: If the word is recognized as subject then Repeat step 8 to 9 for the all the words of the sentence
Step 11: After getting all the words translated, add grammatical suffixes as s, es, ed and t etc with the verb which is necessary.
Step 12: Rearrange the word sequence according to Bengali to English Grammatical rules.
Step 13:-End
Algorithm 2:-
The basic algorithm in distance measuring using the sensor:-
Step 1: "Trigger" pin of the sensor high for 10μs. This initiates a sensor cycle.
Step 2: 8 x 40 kHz pulses will be sent from the transmitting piezzo transducer of the sensor, after which time the "Echo" pin on the sensor will go from low to high.
Step 3: The 40 kHz sound wave will bounce off the nearest object and return to the sensor.
Step 4: When the sensor detects the reflected sound wave, the Echo pin will go low again.
Step 5: The distance between the sensor and the detected object can be calculated based on the length of time the Echo pin is high.
Step 6: If no object is detected, the Echo pin will stay high for 38ms and then go low.
Configuring the 8051 microcontroller:-
To interface the sensor to AT89S51 microcontroller we need two I/O pins. One of them is external interrupt pin (INT0 or INT1) for measuring the pulse width at the echo pin and any other pin say P3.0.
Step 1: Connect the trigger pin of sensor to P3.0 of AT89S51.
Step 2: Connect the echo pin of the sensor to INT0 (P3.1) 0f AT89S51.
Step 3: Configure the TIMER0 of 8051 in 16 bit mode with “GATE” bit enabled. If the GATE pin is enabled and timer run control bit TR0 is set, the TIMER0 will be controlled by INT0 pin. When INT0 is high then the TIMER0 starts counting. Whenever INT0 goes low TIMER0 holds its count. So load the TMOD register with 00001010==0x0A;
Step 4: As the INT0 pin is input don’t forget to declare the pin input. As Write “1” to the pin to make it input.
Algorithm
=>Send a 10 micro second high pulse at trigger
Initially P3.0=0;
P3.0=1;
Delay (10 micro second);
P3.0=0;
=> "WAIT" until the sensor transmits the eight 40 KHz pulses and signal reflection.
Initially the "ECHO" pin is low when the transmitter completes the pulse the pin goes high then our TIMER0 starts counting. When input at INT0 goes low timer holds count.
Logic for waiting:
while (INT0= =0);
while (INT0= =1);
But sometimes due to errors in the sensor functioning the 8051 microcontroller
1) Use watch dog timer (present on AT89S52 and other advanced versions)
2) Generate a delay of 40 milliseconds after triggering the ultrasonic sensor. The second option is preferred for beginners.
=>TIMER0 value = time taken by the signal to (go forward+ come back)
It means the signal traces the whole distance twice. So time taken by the signal to travel the distance = TIMER0 value/2.
X.CONCLUSION
This system also resolves limitations that are related to the most of the movement problems that may influence the blind people in their environment. Future work will be focused on enhancing the performance of the system and reducing the load on the user by replacing the speaker’s tune by real human sound to guide the blind exactly. Moreover, shape detection test for objects that move at different rotational speeds across several distances will further be considered. The advantage of the system lies in the fact that it can prove to be very low cost solution to millions of blind person worldwide. The proposed combination of various working units makes a real-time system that monitors position of the user and provides dual feedback making navigation more safe and secure.