Computer Vision

Zarqa Abid
Introduction

“ To teach computer see and understand , just like we do ”

Introduction...
Computer Vision
Computer Vision...
Computer Vision...
Computer Vision...
Pixel
• Each tiny dot in the picture, called a pixel, is represented by a number
made up of 0s and 1s. This number tells the computer what colour
the pixel should be. The bit plane separates the colour information
into different layers based on these numbers.
• A pixel is the smallest unit of a digital image, representing a single
point of color and intensity.
• DPI vs PPI:
• While PPI refers to the number of pixels per inch in a digital image,
DPI (dots per inch) refers to the number of ink dots a printer can place
in a linear inch, which is used for printing
Pixel...
Screen Size
Intro to Image
Intro to Image...
Pixel...
Pixel...
Color:
• Color in digital images is typically represented using the RGB color
model, where each pixel has values for red, green, and blue, each
ranging from 0 to 255.
• Different combinations of these RGB values create a wide range of
colors.
• For example, (0, 0, 0) represents black, and (255, 255, 255) represents
white.
Pixel...
Intensity (Brightness):
• Intensity, or brightness, refers to how light or dark a pixel appears.
• In grayscale images, a pixel's intensity value directly determines its
brightness, with 0 being black and 255 being white.
• In color images, the intensity is determined by the combined values of the
red, green, and blue channels.

Example:
• If a pixel has RGB values of (255, 0, 0), it would be pure red.
• If a pixel has RGB values of (128, 128, 128), it would be a medium gray.
Color Image
Color Image...
Color Image...
Computer Detect Object
Computer Detect Object...
Binary representation of
integers
• Simplicity: Binary representation simplifies the way computers process
and store information. Since there are only two digits, 0 and 1, it
becomes easier for computers to understand and work with these values.
• Compatibility: Electronics – such as microchips and circuits in computers
and other electronic devices – are designed to work with binary signals.
Binary representation aligns perfectly with the way these devices
function, making it efficient and reliable.
• Compact storage of information: Binary representation allows for
compact storage of information. Since binary digits can be represented by
electrical or magnetic signals, a large amount of data can be stored in a
relatively small space.
• Using the image from the learning hook, investigate what the binary
values are of the RGB values (35,115,166).

• This specific pixel also has a binary value. To find it you need to work
out the binary value of the R, G and B values.To convert a decimal
number to binary, we can divide by 2 and use the remainders to
calculate the binary number.

• Follow these steps:

For the red value of 35

• Start with 35.

• Divide it by 2. The answer is 17 with a remainder of 1.
• Continue dividing: 17 divided by 2 is 8 with a remainder of 1.
• Continue dividing: 8 divided by 2 is 4 with a remainder of 0.
• Continue dividing: 4 divided by 2 is 2 with a remainder of 0.
• Continue dividing: 2 divided by 2 is 1 with a remainder of 0.
• Finally: 1 divided by 2 is 0 with a remainder of 1.
• Write down the remainders in reverse order: 100011.
For the green value of 115
• Start with 115.
• Divide it by 2. The answer is 57 with a remainder of 1.
• Continue dividing: 57 divided by 2 is 28 with a remainder of 1.
• Continue dividing: 28 divided by 2 is 14 with a remainder of 0.
• Continue dividing: 14 divided by 2 is 7 with a remainder of 0.
• Continue dividing: 7 divided by 2 is 3 with a remainder of 1.
• Continue dividing: 3 divided by 2 is 1 with a remainder of 1.
• Finally: 1 divided by 2 is 0 with a remainder of 1.
• Write down the remainders in reverse order: 01110011.
For the blue value of 166

• Start with 166.

• Divide it by 2. The answer is 83 with a remainder of 0.
• Continue dividing: 83 divided by 2 is 41 with a remainder of 1.
• Continue dividing: 41 divided by 2 is 20 with a remainder of 1.
• Continue dividing: 20 divided by 2 is 10 with a remainder of 0.
• Continue dividing: 10 divided by 2 is 5 with a remainder of 0.
• Continue dividing: 5 divided by 2 is 2 with a remainder of 1.
• Continue dividing: 2 divided by 2 is 1 with a remainder of 0.
• Finally: 1 divided by 2 is 0 with a remainder of 1.
• Write down the remainders in reverse order: 10100110.
Results
• So the binary representation of the RGB colour (35, 115, 166) is
(100011, 01110011, 10100110).
Advantages of Computer Vision
 Automation & Efficiency – Computer vision automates tasks like image recognition, object
detection, and quality inspection, reducing human workload.
 High Accuracy & Speed – With advanced AI models, computer vision can process images quickly
and accurately, outperforming human perception in many cases.
 Enhanced Safety – Used in surveillance, autonomous vehicles, and medical diagnostics to prevent
accidents and improve security.
 Scalability – Can analyze large amounts of visual data faster than humans, making it ideal for big
data applications.
 Improved Decision-Making – Helps in medical imaging, defect detection in manufacturing, and
traffic monitoring, leading to better decisions.
 Cost Reduction – Reduces labor costs by automating manual tasks such as monitoring and quality
control.
 24/7 Availability – Unlike humans, AI-driven vision systems work continuously without fatigue.
Disadvantages of Computer
Vision
 High Initial Cost – Developing and deploying computer vision systems require expensive
hardware, software, and skilled personnel.
 Complex Implementation – Requires large datasets, high computational power, and expertise in
machine learning and deep learning.
 Data Privacy Issues – Surveillance and facial recognition systems raise concerns about privacy and
ethical use.
 Limited Generalization – AI models trained for specific tasks may fail in new or unseen conditions
(e.g., different lighting, occlusions).
 Vulnerability to Errors – Misinterpretations in medical diagnosis, self-driving cars, or security
systems can lead to serious consequences.
 Dependence on Data Quality – Requires high-quality and diverse datasets; biased or insufficient
data can result in inaccurate predictions.
 Security Risks – Can be manipulated by adversarial attacks, where small alterations in images trick
AI models into making incorrect decisions.
Uses of computer vision

• 1. Healthcare 🏥
• Medical Imaging – Used in X-rays, MRIs, and CT scans for disease detection (e.g., cancer, fractures,
kidney disease).
• Surgical Assistance – Helps in robotic-assisted surgeries for precision.
• Patient Monitoring – AI-based vision systems track patient movements in hospitals.

• 2. Autonomous Vehicles 🚗
• Object Detection – Identifies pedestrians, traffic signals, and obstacles.
• Lane Detection – Helps self-driving cars stay within lanes.
• Collision Avoidance – Assists in preventing accidents through real-time analysis.

• 3. Manufacturing & Quality Control 🏭

• Defect Detection – Inspects products for flaws in assembly lines.
• Automation – Reduces human intervention in monitoring production processes.
• Safety Compliance – Ensures workers follow safety protocols.
 Uses of computer vision...

• 4. Security & Surveillance 🔍

• Facial Recognition – Used in law enforcement, airport security, and attendance systems.
• Intrusion Detection – Identifies suspicious activities using CCTV footage.
• Number Plate Recognition – Helps in traffic monitoring and law enforcement.

• 5. Retail & E-Commerce 🛒

• Self-Checkout Systems – Uses vision to identify products without barcodes.
• Customer Behavior Analysis – Tracks shopping patterns in stores.
• Virtual Try-On – Allows customers to see how clothes or makeup would look on them.

• 6. Agriculture 🌾
• Crop Health Monitoring – Detects diseases and nutrient deficiencies in plants.
• Automated Harvesting – Helps robots pick fruits and vegetables.
• Livestock Monitoring – Tracks animal movement and health conditions.
 Uses of computer vision...

• 7. Education & Research 📚

• Handwriting Recognition – Used in digitizing handwritten notes.
• Sign Language Recognition – Helps in communication for the hearing impaired.
• Automated Grading – Evaluates handwritten and digital exam papers.

• 8. Defense & Military

• Drone Surveillance – Monitors enemy movements and disaster-affected areas.
• Target Detection – Helps in precision strikes and battlefield analysis.
• Border Security – Detects unauthorized activities at national borders.

• 9. Smart Cities & Traffic Management 🚦

• Traffic Flow Optimization – Uses AI to control signals and reduce congestion.
• Accident Detection – Identifies road incidents and alerts authorities.
• Parking Assistance – Smart systems guide drivers to available parking spots.