Related Works

Reading Text-
Object Obstacle Facial
to-Speech/ ATVIS Description Limitations
Article Detection Detection Recognition
Year OCR
No.
The article introduces a
✔ stereo camera system for  Utilized a bone-conduction
 Stereo Camera obstacle detection using audio technology with
Not Not Not ✔
[69] 2012  Stereo Rig Calibration bone conduction, subpar sound quality and a
Applicable Applicable Applicable
and Rectification validated through a study notable risk of sound
 camera Bumblebee2 with visually impaired leakage.
individuals.

✔
 based on  Balancing accuracy,
Haar The system captures resolution, latency, and
features grayscale face images, power use is challenging,
Not Not Not ✔
[37] 2015  vectors sending them to a server with continuous image
Applicable Applicable Applicable
based on for matching, inspired by processing consuming up to
local binary Casado et al.'s smart 3.5 W, emphasizing energy-
patterns glasses efficient solutions.
(LBP)

The study employs

 The best Obstacle Detection
vehicle obstacle detection
✔ technique will be
for visually impaired
incorporated into an
individuals, using
 Ultrasonic Not Not ✔ optimized system, with
[75] 2015 Not ultrasonic sensors and
Sensor Applicable Applicable future plans for a
Applicable smartphone-based
 Arduino Board recognition system in a
computer vision,
 OpenCV navigation tool for the
enhancing collision-free
visually impaired.
mobility and navigation.

The paper presents a live-

✔  The system recognizes over
time face recognition
 Epson 85% of faces in real-time
system for smart
Not Not Not Moverio ✔ within 0.3m to 0.8m,
[38] 2016 eyeglasses, assessing its
Applicable Applicable Applicable (BT-200) peaking at 98% accuracy at
performance on Epson's
 Info-Linker 0.5m. Accuracy declines
Moverio and InfoLinker
WUZ01A beyond 1.1m.
platforms.
✔
 Maximally The paper proposes a
Stable camera system for
 The system lacks audio
External visually impaired text
✔ feedback for obstacles,
Not Regions Not ✔ reading, employing
[50] 2016  Camera is of emphasizing a primary focus
Applicable (MSER), Applicable MSER for detection and
5MP on text detection and
Optical OCR for speech, yielding
recognition in the research.
Character positive experimental
Recognition outcomes.
(OCR)
The paper suggests a
✔ real-time object detection
✔
 Microsoft system for the visually  Limited for indoor use only
 Camera Module Not Not ✔
[59] 2016 Kinect impaired, utilizing and no direct sunlight for the
 Using Depth Image Applicable Applicable
 Camera Microsoft Kinect with camera module.
System
Module 92% accuracy in blind
 Checkmark (✔) – Applicable
 ATVIS – Applicability to Visually Impaired Students