There are a couple of million visually challenged people over the globe. Vision is considered to be one of the most essential of human senses and plays vital role in our life. In the modernized world, lot of people with disabilities are capable of entering in the work force and competing with ordinary people. The reservation and priority offered to them by the government help encourage their empowerment. The government also provides subsidies for getting walking aids, hearing aids, Braille, etc. There is a high possibility that government may provide these technologically developed equipment’s completely free of cost or at low cost. The equipment’s available in the market are helping the visually disabled and are much useful but they too have limitations. Considering this limitation, a need to develop a new equipment with innovative technology was proposed. The novel Retinal Art is designed to satisfy the limitations and meet the demands of the visually disabled. The venture for the most part focuses on visually disabled people whose optical nerves are still undamaged.
According to the anatomy of the brain, the optical nerves, in charge of sight and the glossopharyngeal and carnival nerve in charge of taste cover one another in the neo-cortex locale of the cerebrum. By utilizing this covering of nerves as leverage, fitting application of the signals on the tongue can be anticipated by the users which permits them to see the objects around them. The picture caught from the webcam is converted to grey scale image of 64 pixels i.e.an electrode of 8x8. This gadget is interfaced to the image processing unit through the sound ports and the output of the image processing unit is processed and given as an input to the tongue of the impeded individuals.
Structure of the Interface
The primary focus of the work is to make the visually impaired individuals to perceive the items around them. This system permits gathering the information from a webcam and sends the signal for imagining it on a viewer array set on a singular's tongue. The task is to make an 8x8 tangible lattice, 64 pixels, utilizing grey scale with a 255 resolution. From a specialized perspective, the picture is not static and has a redundancy recurrence of around 50 Hz. This recurrence similar as in case of television, is focused around the way that at this reiteration speed the cerebrum is unequipped for detecting the picture. Hence the picture is drawn pixel by pixel as an edge with 64 pixels, at a rate of 50 Hz every second. In this work, the data is utilized such that the response speed of a neuron is p200, which demonstrates that a neuron can't be enacted in less than 200 milliseconds.
Device Structure
Camera
Digital cameras capture images as pixel element, known as pixels. Simply put, a Mega pixel is equal to one million pixels. Digital images are made up of thousands of these tiny, tile-like picture elements .The more pixels, the higher the image resolution. Resolution relates primarily to print size and the amount of detail an image has when viewed on a computer monitor at 100%. Single shot capture systems use either one sensor chip with a Bayer filter mosaic, or three separate image sensors (one each for the primary additive colours red, green, and blue) which are exposed to the same image via a beam splitter.In this project 2MP camera is used. The Camera is placed such that it can rotate in all the directions. This facilitates the visually challenged people to see in all the directions without turning actually. Else, a three button sized cameras are placed in the spectacles of the person which will be switched on when required by the impaired person.
Image processing unit
The picture taken from the webcam is handled by programming and modified with the help of Artificial intelligence to perceive human faces. Signal Generator is a software that permits facial distinguishment focused around multi-layer recognition systems. After facial recognition, the product can be utilized for facial distinguishment, utilizing a cascaded neural system.This product incorporates the procurement of distinctive sorts of signs from the caught picture. As the gadget is interfaced with machine through the sound port, right and left volume adjustments are given. The pictures caught in different locales will have different noise. For extraction of the content, the clamour is uprooted. To evacuate the commotion and to make modification in pictures, different tools are utilized. With a specific goal to apply the important sign to tongue, steps are to be emulated.
• The captured picture is changed over to grey scale utilizing grey scale transformation strategy.
• The contrast enhancements can be done in histogram adjustment tool.
• To kill the clamours, convolution channels and white Gaussian channels are utilized.
• For content extraction and to get the distinction in different levels between the pixels the tools like Laplace, prewitt, sobel, enhance are utilized.
• To include the quantity of items in the picture, segmentation is utilized.
• Finally the background is uprooted.
• The obliged signs will be sent to the hardware components as an audio signal on two sound channels: one channel with the substance of the picture and with a frame separator; and the other channel just with a frame separator for synchronization.
Comparator
TLC277 consists of two operational amplifiers. These operational amplifiers are configured to work as a comparator. Since the op amp is capable of operating with the analogical inputs, the audio from the image processing unit which is analog in nature can be compared through an op-amp circuitry built in TLC277.The circuit amplifies the voltage difference between Vin and VRef, and outputs the result at Vout. If Vin is greater than VREF, then voltage at Vout will rise to its positive saturation level; that is, to the voltage at the positive side. If Vin is lower than VREF, then Vout, will fall to its negative saturation level, equal to the voltage at the negative side. In practice, this circuit can be improved by incorporating a hysteresis voltage range to reduce its sensitivity to noise. The positive voltage obtained as output from comparator can be considered to be logic 1 and the negative voltage will be logic 0.If the frame separator from both the left and right ports are synchronized, it produces an high output which enables the multi-vibrator.Multivibrator
The high output from the comparator enables the multivibrator to send the clock pulses to the counter. The comparator output pulses acts as an enabling pulse for multivibrator. The multivibrator, CD4047 is configured to work in astable mode. It is made to work in astable mode by connecting a capacitor between the pin 1 and 3 and a resistor between the pin 2 and 3 for the proper working of IC. The output obtained at pin 11 enables the counter.Counter
The square wave output from the multivibrator acts as the clock pulse for the counter. For each square wave pulse, the counter increases the count by one, starting from zero to sixty three. Thus totally it makes 64 counts for each image frame. After attaining the maximum count, the counter starts counting in the reverse direction for the next image frame. These counts help the demultiplexer to demultiplex the signals on the 64 electrode in the array.Demultiplexer
The CD 4067B a demultiplexer from Texas Instruments is a 1:16 demultiplexer. The pin C I/O acts as input and the pins 2 to 9 and 16 to 23 acts as output. To build a 1:64 demultiplexer unit, five 1:16 demultiplexers are required, one for sending the enabling pulse (i.e. it acts a master) and the remaining four for demultiplexing the signal from the operational amplifier. Thus 64 electrodes in the viewer array are energized one after the other, causing the tingling sensation in the tongue.Viewer Array
Viewer array consists of an array of electrodes. It is placed on the tongue of the blind person to have a contrast access to the external environment. The output signal applied to the viewer array is multiplied with the help of CD4067B de-multiplexer. The electrodes in the viewer array is made of aluminium. Aluminium is remarkable for the metal’s low density and for its ability to resist corrosion due to the phenomenon of passivation. Aluminium being the good conductors, durable light weight, ductile, malleable metal, corrosion resistance and cheap it is used to build the array.Hence the device should generate the current between 1 and 5mA to make a tingling sensation to teach blind people regarding the shape and movement of the articles around them. The current less than 1mA is imperceptible hence it should be above 1mA and should not exceed 5mA. The output current tested from array of electrodes is limited such that it lies between 1 and 5mA. Thus with the help of tingling sensation caused on the tongue, the blind people will learn the shape of different objects and their movement which makes them to experience the real sense of sight.
Experimental Results
Signal Generator software is capable of detecting the face of humans, recognize letters and objects in front of them. The picture taken from the webcam is handled by programming and modified with the help of Artificial intelligence to perceive human faces. This software produces audio signal as the output which can be interfaced to equipment through the sound port. The required image processing tools to eliminate the clamour and the tools for extracting the desired content from the images, is built-in under the filter section in the software. These tools when clicked appears at the graph area where the appropriate connection is made, then the simulation is started.The pictures caught in different locales will have different noise. For extraction of the content, the clamor is uprooted. To evacuate the commotion and to make modification in pictures, different tools are utilized. The generated array displays the electrodes which are energized in the viewer array and the signals are sent to the viewer array through the audio channel. When the viewer array is placed in the tongue the respective electrodes which are energized causes a tingling sensation in the tongue of the visually impaired person. With the help of this, the objects around them can be detected. Thus, they can have continuous access to the external environment.
This is the project carried out by my team in final year of Bachelor of Engineering. If you like this project please kindly support us through your comment and likes.
Comments
Post a Comment