Top Banner
Presented by, VINOTHINI.N SORNA LAKSHMI.S For Aiding Navigation Of Visually Impaired Persons
13
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: An embedded system for aiding navigation of visually

Presented by,VINOTHINI.N

SORNA LAKSHMI.S

An Embedded System For Aiding Navigation Of Visually Impaired

Persons

Page 2: An embedded system for aiding navigation of visually

OBJECTIVE• Visually impaired individuals find navigation difficult as they often lack the much needed

information for bypassing obstacles and hazards in their path.

Page 3: An embedded system for aiding navigation of visually

CONTENT

Introduction

Proposed electronic travel aid

Ultrasound-based distance measurement

Detection of human presence

Face detection

Detection of cloth

Result and detection

Conclusion

Page 4: An embedded system for aiding navigation of visually

INTRODUTION• Electronic travel aid (ETA) is one of the new approaches to the solution,allowing the visually

impaired to live independent life.

• A headphone is connected with eBox 2300™ to get the beep sound.

• They used a digital video camera as video sensor…

• A USB webcam is connected with ebox2300.

• Embedded System for capturing the field-of-view.

• Face detection technique.

• Cloth texture analysis small image frame (160 × 120) to achieve real-time image-processing.

• X86 processor-based embedded computer system15.

• It feeds the distance data to eBox 2300™ through a RS-232 serial cable

Page 5: An embedded system for aiding navigation of visually

DETECTION OF HUMAN PRESENCE

• Human presence is detected by the human face.

• However, there are situations when the face is not present in the field-of-view of the camera in spite of the presence of a human being in front of the visually impaired person.

• Such type of presence of human beings may be asserted by detecting cloth and human skin.

• To detect the presence of humans, if cloth is found in the vicinity of human skin and face is not detected (side faces), then it will be considered as human.

Page 6: An embedded system for aiding navigation of visually

ULTRASOUND-BASED DISTANCE MEASUREMENT• The overall operation is divided into two parts:

o development of sensor unit.

o processing of sensor circuit data.

TRANSMITTER(40KHZ)>> RECEIVER>>PROCESSING UNIT(DISTANCE)

• Distance in the form of pulse count as its output of sensor unit.

• These counts are used to compute actual distance through processing in eBox 2300 through RS232 cable.

• The eBox 2300™ generates a temporary file

• A C++ program running on eBox 2300™ reads this file to process the data.

• The system detects obstacles up to 300 cm.

• Finally Beep sound is provided to the user.

Page 7: An embedded system for aiding navigation of visually

DETECTION OF FACE• In the Segmentation of skin region,the input RGB

image is transformed to YCrCb colour space and the skin pixels are detected based on their chrominance values.

• 240 frontal face images of size 160 × 120 of 12 individuals have been taken at distances 30, 60, 90 and 120 cm.

• The USB webcam connected with eBox 2300™ embedded system provides image frames at a rate of 15 fps,

• In the non-face image though false skin region has been segmented out, its feature values do not satisfy all three conditions (Fbe = 1.48, Fab = 1.28 and Foe = 1.188).

• Hence it is classified as non-face.

Page 8: An embedded system for aiding navigation of visually

DETECTION OF CLOTH• The image is subdivided into 48 non- overlapping sub images of size 20 × 20 pixels

• The energy values of eA, eH, eV, eD .

• The feature values are summarized into mean ( μ ) and standard deviation ( σ ) and the 3 σ -control limits are taken

• The probability of having values in the interval ( μ – 3 σ , μ + 3 σ ) is very large (nearly 0.9973).

• The decision is made whether the texture is cloth or not.

• The input image has different cloth textures at 120 cm distance.

Page 9: An embedded system for aiding navigation of visually
Page 10: An embedded system for aiding navigation of visually

RESULT

• The USB webcam is mounted on a helmet and ultrasonic sensors are placed on the user’s belt.

• Sensor circuit unit and eBox 2300™ are carried by the user in a bag..

• Three easy control switches: First switch- To find the obstacles in the path second -To find human presence in the field-of-view

Third switch- Detect any movement in front of the person.

• If the user presses the first button, the device will produce a beep sound

• The accuracy of the device in finding obstacles is 95.45%.

Page 11: An embedded system for aiding navigation of visually

CONCLUSION

• An ETA to navigate visually impaired persons has been proposed here.

• UThe major issues for the users to accept these aids are that they should be unobtrusive, easy to carry, small and light weight. .

• The user needs to wear a helmet in which the camera and headphone are mounted.

• The user has to carry eBox 2300™ and sensor unit of nearly 500 g.

• This device can detect obstacles up to 300 cm and human presence within 120 cm.

• Thus, a portable ETA device has been developed taking into account the blind users’ requirements and they can pass through unknown environment independently..

Page 12: An embedded system for aiding navigation of visually

Queries

?

Page 13: An embedded system for aiding navigation of visually

Thank you