Lecture-10
Computer Vision
By : Dr. Shimaa Saber
1
Course Ethics
Agenda
➢Stereo Vision
1. Introduction
2. Algorithms
• Depth
3. applications
3
Stereo Vision
Introduction
4
Stereo Vision
Stereo vision, also known as stereoscopic vision or stereopsis.
Stereo Vision refers to the ability of an organism to perceive
depth or three-dimensional (3D) information by processing
visual stimuli from both eyes.
Humans and many animals possess stereo vision, and it plays
a crucial role in depth perception, which is the ability to
judge the distance of objects in the environment.
5
Introduction
6
Introduction
Computer stereo vision is the extraction of 3D information
from digital images. By comparing information about a scene
from two vantage points, 3D information can be extracted by
examination of the relative positions of objects in the two
panels.
➢ Two Seeing Eyes = Two Views!
➢ Two Views Used and Fused in the Brain = Stereovision!
7
Introduction
Stereo vision refers to the use of two or more cameras to
simulate human-like depth perception.
This technique is based on the same principles of stereopsis
used by humans. Capturing images of a scene from different
viewpoints
Computer systems can analyze the disparities in the images
to deduce the depth information, creating a three-
dimensional (3D) representation of the scene.
8
Introduction
Structure from stereo (Stereo-vision): use two cameras with
known relative positions and orientation
9
Stereo Vision Algorithms
Simplest Case
10
Stereo Vision- Simplified Case
The simplified case
is ideal. It assumes
that both cameras
are identical and
are aligned on a
horizontal axis
𝑅1 , 𝑡1 Camera 2
11 Camera 1
Stereo Vision- Simplified Case
The simplified case assumptions
12
Stereo Vision- Simplified Case
Baseline: distance between
the optical centers of the
two cameras
13
Introduction
disparity: the difference in image location of the same 3D point when
projected under perspective to two different cameras
14
Stereo Vision- Simplified Case
15
Stereo Vision- Simplified Case
f = focal
b = baseline
z = depth
d = disparity
d = 𝑥𝐿 - 𝑥𝑅 𝑍𝐿 = 𝑍𝑅 = Z 𝑋𝐿 = 𝑋𝑅 + b
𝑋𝐿
𝑥𝐿 = 𝑓 ∗ 𝑋𝑅 +b 𝑋𝑅
𝑍𝐿 d = 𝑥𝐿 - 𝑥𝑅 = 𝑓 ∗ -𝑓 ∗
𝑋𝑅 𝑍 𝑍
𝑥𝑅 = 𝑓 ∗ 𝑓 ∗ 𝑏 𝑓 ∗ 𝑏
𝑍𝑅 𝑑= 𝑍=
16 𝑧 𝑑
Stereo Vision- Simplified Case
A stereo vision system estimates the disparity of a point as
d=10 pixels. What is the depth (Z) of the point, if f = 500
pixels and b = 10 cm?
f = 500 pixel
b = 10 cm 𝑓∗𝑏 500 𝑝𝑖𝑥𝑒𝑙∗10 𝑐𝑚
𝑍= =
z =? 𝑑 10 𝑝𝑖𝑥𝑒𝑙
d = 10 pixels = 500 𝑐𝑚 = 5 𝑚
17
Stereo Vision- Simplified Case
Find a 3D point corresponding to 2 points P1 and P2 in
from the right and left camera respectively,
where P1 (88,90), P2 (100,90). f=500 cm, b=10 pixel.
18
Stereo Vision
Applications
19
Applications
Drone Navigation
20
Applications
Robotics: for navigation, obstacle avoidance, and
object manipulation.
Augmented Reality: It enhances the realism of
augmented reality applications by accurately placing
virtual objects in the real world.
3D Mapping: Stereo vision helps create detailed 3D
maps of environments for applications like
autonomous vehicles and virtual reality.
Medical Imaging: for tasks such as reconstructing 3D
21
structures from multiple images.
Summary
➢Stereo Vision
1. Introduction
2. Algorithms
• Depth
3. applications
22
23
