UNIT–1
Overview of Computer Vision and Image
Processing: Definitions and Scope
Vision
Vision is the ability to perceive, identify, and understand objects and the surrounding
environment through light entering the eyes.
Types
1. Human Vision
2. Computer Vision
Real-World Examples
A person recognizing a friend's face.
A driver seeing traffic signals.
A student reading a book.
Diagram
Light
↓
Human Eye
↓
Brain
↓
Understand Scene
Applications
Reading books
Driving vehicles
Identifying objects
Recognizing faces
Human Vision
Human Vision is the natural ability of humans to observe, recognize, and interpret the
surrounding environment using the eyes and brain.
Components
Eye
Retina
Optic Nerve
Brain
Work flow
Object
↓
Light Reflection
↓
Eye
↓
Retina
↓
Optic Nerve
↓
Brain
↓
Recognized Object
Real-World Example
You see a mango on a table and immediately recognize it because your brain has learned its
shape and color.
Advantages
Learns from experience
Adapts to different environments
Excellent object recognition
Limitations
Gets tired
Affected by lighting conditions
Subject to human error
Computer Vision
Computer Vision is a branch of Artificial Intelligence (AI) that enables computers to acquire,
process, analyze, and understand images or videos to make decisions automatically.
Types (Major Tasks)
Image Classification
Object Detection
Image Segmentation
Face Recognition
Image Tracking
Computer Vision Workflow
Real World Object
│
▼
Image Acquisition
(Camera / Sensor)
│
▼
Image Preprocessing
(Resize, Filter, Threshold)
│
▼
Feature Extraction
│
▼
Machine Learning / AI Model
│
▼
Object Recognition / Classification
│
▼
Decision / Output
Real-World Examples
Face Unlock on smartphones
Self-driving cars detecting pedestrians
CCTV detecting intruders
Google Lens identifying objects
Medical diagnosis from X-rays
Advantages
High speed
High accuracy
Works continuously without fatigue
Processes large volumes of images
Limitations
Requires quality data
Computationally intensive
Sensitive to poor lighting and image quality
Image Processing
Definition
Image Processing is the process of performing operations on a digital image to improve its
quality or extract useful information.
Types
Based on Processing
1. Analog Image Processing
2. Digital Image Processing
Common Operations
Brightness Adjustment
Contrast Adjustment
Filtering
Thresholding
Histogram Processing
Image Processing Workflow
Input Image
│
▼
Image Enhancement
│
▼
Noise Removal
│
▼
Feature Extraction
│
▼
Output Image
Scope of Computer Vision
Definition
The scope of Computer Vision refers to the wide range of fields where image analysis and visual
understanding techniques are applied to solve real-world problems.
Applications of Computer Vision
Applications of Computer Vision are the real-world areas where computers analyze images or
videos to perform useful tasks automatically.
Major Application Area
1. Healthcare
MRI Scan Analysis
CT Scan Analysis
Tumor Detection
Disease Diagnosis
Example:
AI detecting lung cancer from X-ray images.
2. Security and Surveillance
Face Recognition
CCTV Monitoring
Intruder Detection
Example:
Airport security systems.
3. Autonomous Vehicles
Lane Detection
Traffic Sign Recognition
Pedestrian Detection
Obstacle Detection
Example:
Self-driving cars.
4. Agriculture
Crop Disease Detection
Fruit Quality Inspection
Weed Detection
Example:
Drone monitoring crop health.
5. Manufacturing
Defect Detection
Product Quality Inspection
Robot Guidance
Diagram
Computer Vision
│
┌──────────┬──────────┬──────────┬──────────┐
│ │ │ │ │
Healthcare Security Robotics Agriculture Transport
Real-World Examples
Healthcare
Detecting tumors from MRI and CT scan images.
Security
Face recognition at airports.
Agriculture
Identifying plant diseases from leaf images.
Transportation
Automatic number plate recognition.
Industry
Detecting defective products on manufacturing lines.
Advantages of Computer Vision
Reduces human effort
Fast image analysis
Improves accuracy
Works 24×7
Supports automation
Limitations
Needs high-quality images
High computational cost
Performance decreases with poor lighting
Requires training data for AI-based systems
Historical Development
The historical development of Computer Vision and Image Processing refers to the evolution
of technologies, techniques, and algorithms that enable computers to acquire, process, analyze,
and understand images from the real world.
Timeline of Historical Development
Year Development Description
Digital Image
1950s Scientists began processing digital images using computers.
Processing
NASA used image processing to enhance images from satellites
1960s Space Image Processing
and spacecraft.
Medical Image
1970s CT scans and MRI imaging became popular in healthcare.
Processing
1980s Pattern Recognition Computers started recognizing shapes, numbers, and characters.
Early face detection and biometric security systems were
1990s Face Recognition
developed.
Computer vision systems learned from training data to improve
2000s Machine Learning
recognition.
Year Development Description
CNNs greatly improved image classification and object detection
2010s Deep Learning
accuracy.
AI-Based Vision Used in autonomous vehicles, smart surveillance, robotics, and
2020s
Systems healthcare.
Timeline Diagram
1950s → Digital Image Processing
↓
1960s → Satellite Image Processing
↓
1970s → Medical Imaging
↓
1980s → Pattern Recognition
↓
1990s → Face Recognition
↓
2000s → Machine Learning
↓
2010s → Deep Learning
↓
2020s → AI-Based Computer Vision
Evolution of Computer Vision
Image Capture
↓
Image Processing
↓
Pattern Recognition
↓
Machine Learning
↓
Deep Learning
↓
Artificial Intelligence
↓
Smart Computer Vision Systems
Advantages
Fast image analysis
High accuracy
Reduces manual effort
24×7 operation
Supports automation
Improves decision making
Limitations
Requires high-quality images
High computational requirements
Performance affected by poor lighting
Requires trained AI models and datasets
Image Formation
Image Formation is the process by which light reflected from an object passes through an optical
system (camera lens or human eye) and forms an image on a sensor or retina.
Types
1. Binary Image
Pixel values: 0 or 1
Black and White only
Example:
QR Code
2. Grayscale Image
Pixel values: 0–255
Black to White
Example:
X-ray image
3. Color Image
Three color channels
o Red
o Green
o Blue
Example:
Mobile phone photograph
Image Formation Process (Workflow)
Real World Object
│
▼
Light Reflection
│
▼
Camera Lens / Human Eye
│
▼
Image Sensor / Retina
│
▼
Digital Image
│
▼
Computer Storage
Image Representation
Image Representation is the process of storing an image digitally as a collection of pixels, where
each pixel contains intensity or color information.
Inside a computer, an image is stored as rows and columns of pixels.
Example:
Image
┌──┬──┬──┬──┐
│25│80│60│40│
├──┼──┼──┼──┤
│55│90│75│20│
├──┼──┼──┼──┤
│15│45│65│95│
└──┴──┴──┴──┘
Each number represents the intensity value of one pixel.
Real-World Examples
Mobile Camera
The mobile camera captures your selfie and stores it as millions of pixels.
CCTV Camera
Captures continuous video frames.
Each frame is an image.
Satellite Imaging
Earth images captured by satellites are represented as digital pixels.
Medical Imaging
MRI and CT scans are stored as grayscale images.
Applications
Digital Photography
Medical Imaging
Advantages
Easy to store
Easy to process
High accuracy
Supports AI applications
Enables image enhancement
Limitations
Large images require more storage.
High-resolution images require more processing time.
Image quality depends on camera resolution and lighting conditions.
Image Acquisition
Image Acquisition is the process of capturing an image from the real world using an imaging
device (camera or sensor) and converting it into a digital image that can be processed by a
computer.
Why is Image Acquisition Important?
Without image acquisition, a computer has no visual data to process.
It is the starting point for all Computer Vision tasks such as:
Face Recognition
Medical Imaging
Object Detection
OCR
Satellite Image Analysis
Image Acquisition Workflow
Real World Object
│
▼
Light Falls on Object
│
▼
Camera / Sensor
│
▼
Analog Signal
│
▼
Analog-to-Digital Converter (ADC)
│
▼
Digital Image
│
▼
Computer Vision Processing
Image Acquisition Methods
There are several methods to acquire images.
Method 1: Camera-Based Acquisition
Definition
A digital camera captures light reflected from an object and converts it into a digital image.
Examples
Mobile Camera
DSLR Camera
Webcam
CCTV Camera
Applications
Photography
Face Recognition
Video Conferencing
Surveillance
Method 2: Scanner-Based Acquisition
Definition
A scanner captures printed documents or photographs and converts them into digital images.
Examples
Flatbed Scanner
Document Scanner
Barcode Scanner
Applications
Document Digitization
OCR
Banking
Libraries
Method 3: Medical Imaging Acquisition
6
Definition
Medical devices capture images of internal body structures for diagnosis.
Examples
MRI
CT Scan
X-ray
Ultrasound
Applications
Tumor Detection
Bone Fracture Analysis
Organ Examination
Method 4: Satellite and Remote Sensing
6
Definition
Satellites capture images of the Earth's surface from space.
Applications
Weather Forecasting
Agriculture
Disaster Management
Urban Planning
Method 5: Industrial Image Acquisition
6
Definition
Industrial cameras capture images during manufacturing for inspection and automation.
Applications
Quality Inspection
Defect Detection
Robotics
Automated Assembly
5. Comparison of Image Acquisition Methods
Method Device Applications
Camera-Based Mobile, DSLR, Webcam Photography, Face Recognition
Scanner-Based Flatbed Scanner OCR, Document Scanning
Medical Imaging MRI, CT, X-ray Healthcare
Satellite Imaging Satellites Agriculture, Mapping, Weather
Industrial Imaging Machine Vision Cameras Manufacturing, Inspection
6. Real-World Examples
Mobile Phone
Take a selfie → Camera captures the image → Stored digitally.
ATM
Cheque scanner reads a cheque and converts it into a digital image.
Hospital
CT Scanner captures internal body images.
Google Maps
Satellites capture Earth's surface images.
Factory
Machine vision cameras inspect products for defects.
7. Applications
Computer Vision
Medical Diagnosis
OCR
Autonomous Vehicles
Robotics
Smart Agriculture
Traffic Monitoring
Security Surveillance
Industrial Automation
8. Advantages
Captures real-world information digitally.
Enables automated image analysis.
High-speed acquisition.
Supports AI and Machine Learning applications.
9. Limitations
Poor lighting affects image quality.
Expensive sensors may be required.
Noise can reduce image quality.
Weather conditions can affect outdoor image capture.