COMPUTER

VISION AND
IMAGE
PROCESSING
Dr.S.Nagarajan
AGENDA
What is Computer Vision?
What is Image?
Related Fields in Computer Vision
CV vs Image Processing
Applications of CV and IP
Different Image Processing Examples
Fundamental Steps in Image Processing
 3

What is Computer Vision?

Computer vision is a field of Artificial Intelligence (AI) that

enables computers and systems to derive meaningful
information from digital images, videos and other visual
inputs — and take actions or make recommendations based on
that information. If AI enables computers to think, computer
vision enables them to see, observe and understand.

Source : https://www.ibm.com/topics/computer-vision
HOW DOES COMPUTER VISION WORK? 4
 5

HOW DOES COMPUTER VISION WORK?

Computer vision mainly performs three steps, which are:

1. Capturing an Image :
A computer vision software or application always includes a digital camera or
CCTV to capture the image. So, firstly it captures the image and puts it as a
digital file that consists of Zero and one's.
2. Processing the image
In the next step, different CV algorithms are used to process the digital data
stored in a file. These algorithms determine the basic geometric elements and
generate the image using the stored digital data.
3. Analyzing and taking required action
Finally, the CV analyses the data, and according to this analysis, the system
takes the required action for which it is designed.
 6

WHAT IS IMAGE?
An image is defined as a two-dimensional function(x,y),
where x and y are spatial(plane) coordinates, and the
amplitude of F at any pair of coordinates (x,y) is called the
intensity of that image at that point. When x,y, and
amplitude values of F are finite, we call it a digital image.

Gray level is used often to refer to the intensity of

monochrome images.

Color images are formed by combination of individual

 7

• RELATED FIELDS IN COMPUTER

• Hardware
Machine Learning and Deep Learning
Development VISION
• Image Processing
• Pattern Recognition
• Robotics
• Natural Language Processing (NLP)
• Biometrics
• Augmented Reality (AR) and Virtual Reality
(VR)
• Medical Imaging
• Autonomous Vehicles
• Remote Sensing
• Agriculture
• Security and Surveillance
• Entertainment and Gaming
• Geographic Information Systems (GIS)
• Artificial Intelligence Ethics and Bias
 8

RELATED FIELDS IN COMPUTER VISION

1. Machine Learning and Deep Learning: Machine learning and deep learning are essential
components of computer vision. Techniques such as convolutional neural networks (CNNs)
and Recurrent Neural Networks (RNNs) are widely used for image and video analysis.

2. Image Processing: Image processing deals with manipulating and enhancing images to
improve their quality or extract useful information. Many computer vision applications
involve image processing techniques.

3. Pattern Recognition: Pattern recognition is the study of recognizing patterns within data. In
computer vision, this involves recognizing objects, shapes, or gestures in images or videos.

4. Robotics: Computer vision plays a significant role in robotics, enabling robots to perceive
and navigate their environment. It's used in tasks like object detection, localization, and
mapping (SLAM). (simultaneous localization and mapping)
 9

RELATED FIELDS IN COMPUTER VISION

5. Natural Language Processing (NLP): Combining computer vision with NLP
allows for multimodal analysis, where both text and visual data are used together,
enabling applications like image captioning and sentiment analysis of images.
6. Biometrics: Biometrics involves recognizing individuals based on their unique
physical or behavioral characteristics. Face recognition and fingerprint analysis
are examples of computer vision applications in biometrics.
7. Augmented Reality (AR) and Virtual Reality (VR): Computer vision is essential in
AR and VR for tracking the user's movements and the environment to provide an
immersive experience.
8. Medical Imaging: In the field of medical imaging, computer vision is used for
tasks like image segmentation, tumor detection, and disease diagnosis through the
analysis of medical images like X-rays, MRIs, and CT scans.
 10

RELATED FIELDS IN COMPUTER VISION

9. Autonomous Vehicles: Computer vision is a key technology in the

development of self-driving cars. It's used for tasks like lane detection, object
detection, and path planning.
10.Remote Sensing: Remote sensing involves collecting data about the Earth's
surface from a distance, often using satellite imagery. Computer vision is used
for land cover classification, monitoring natural disasters, and more.
11.Agriculture: In precision agriculture, computer vision is used for crop
monitoring, yield prediction, and disease detection in plants.
12.Security and Surveillance: Computer vision is used for video surveillance,
facial recognition, and anomaly detection in security applications.
 11

RELATED FIELDS IN COMPUTER VISION

13. Entertainment and Gaming: The gaming industry often uses computer vision
for motion capture, gesture recognition, and creating immersive experiences.
14.Geographic Information Systems (GIS): Computer vision is used to extract
geographical information from images and satellite data, assisting in map
creation and geospatial analysis.
15.Artificial Intelligence Ethics and Bias: As computer vision systems become more
pervasive, ethical considerations related to privacy, bias, and fairness in AI are
gaining importance.
16.Hardware Development: The development of specialized hardware, such as
GPUs and TPUs, plays a crucial role in accelerating the computation required
for computer vision tasks.
 12

COMPUTER VISION VS IMAGE PROCESSING

Computer Vision:
1. Goal: Computer vision focuses on enabling computers to interpret and understand the visual
world, much like human vision. It's about replicating human vision and cognition in machines.
2. Tasks: Computer vision involves higher-level tasks such as object recognition, scene
understanding, image understanding, and even making decisions based on visual input. This can
include tasks like object detection, image segmentation, facial recognition, and 3D reconstruction.
3. Complexity: Computer vision tasks often involve understanding the context and semantics of
images. It requires higher-level reasoning and can involve machine learning and AI techniques,
including deep learning.
4. Applications: Computer vision is used in various applications, including autonomous vehicles,
robotics, medical image analysis, and augmented reality. It often deals with real-world, dynamic
environments.
5. Example: An autonomous car using computer vision might identify pedestrians, other vehicles,
traffic signs, and road conditions to make driving decisions.
 13

COMPUTER VISION VS IMAGE PROCESSING

Image Processing
1. Goal: Image processing is primarily focused on manipulating and enhancing images to extract
information or improve image quality. It deals with the modification of the visual content itself.
2. Tasks: Image processing tasks include tasks like image filtering, noise reduction, image
sharpening, contrast enhancement, and color correction. It's often concerned with low-level
operations on images.
3. Complexity: Image processing typically involves simpler operations that are applied to
individual pixels or small neighborhoods of pixels within an image. It's not concerned with
high-level interpretation or recognition.
4. Applications: Image processing is used in applications like image compression, medical image
enhancement, and improving the quality of photos. It's commonly applied in scenarios where
the goal is to improve the visual quality or extract specific information from images.
5. Example: A medical image processing application might enhance the contrast in an X-ray
image to make it easier for a radiologist to identify abnormalities.
 14

APPLICATIONS OF COMPUTER VISION AND IMAGE PROCESSING

1. Object Recognition and Detection: Computer vision is used to recognize and locate objects in
images or video streams, such as pedestrian detection in autonomous vehicles, facial
recognition for security, and barcode scanning in retail.
2. Image Segmentation: This involves dividing an image into meaningful parts, which is used in
medical image analysis, industrial quality control, and scene understanding in robotics.
3. Gesture Recognition: Computer vision enables the interpretation of hand and body gestures,
which is useful in applications like sign language translation, gaming, and human-computer
interaction.
4. Face Recognition: Facial recognition is widely used in security, access control, and
authentication systems, as well as in social media for tagging photos.
5. Optical Character Recognition (OCR): OCR technology is used to convert printed or
handwritten text in images into machine-readable text. Applications include digitizing printed
documents, license plate recognition, and reading text from images for the visually impaired.
 15

6. Robotics: Computer vision plays a crucial role in enabling robots to perceive their environment and
navigate, allowing them to perform tasks like object manipulation and autonomous navigation.
7. Augmented Reality (AR) and Virtual Reality (VR): AR and VR applications use computer vision for
tracking users' movements and the environment, enabling immersive experiences and overlaying
digital information on the real world.
8. Medical Image Analysis: Computer vision is used in medical imaging to analyze X-rays, MRIs, CT
scans, and other medical images for tasks like tumor detection, organ segmentation, and disease
diagnosis.
9. Autonomous Vehicles: Self-driving cars rely on computer vision for tasks like lane detection, obstacle
detection, and road sign recognition.
10. Agriculture: Computer vision is used for crop monitoring, yield prediction, and disease detection in
precision agriculture.
11. Industrial Quality Control: Computer vision systems are used to inspect and ensure the quality of
manufactured products, such as checking for defects in electronics, textiles, and food items .
 APPLICATIONS OF IMAGE 16

PROCESSING:
1. Image Enhancement: Image processing techniques are used to enhance the
quality of images, improve contrast, reduce noise, and adjust brightness and
color levels.
2. Image Compression: Compression algorithms are used to reduce the size of
images for efficient storage and transmission, as in JPEG and PNG formats.
3. Image Restoration: Restoring old or damaged images by removing scratches,
dust, or other imperfections is a common application.
4. Image Filtering: Techniques like edge detection and blurring are used for
various applications, including feature extraction and special effects in
photography.
5. Color Correction: Adjusting the color balance, saturation, and contrast of
images to improve their visual quality in photography and video editing.
 APPLICATIONS OF IMAGE 17

PROCESSING
6. Morphological Operations: These are used in applications like document
processing and shape analysis to extract or manipulate image features.
7. Image Registration: Aligning multiple images of the same scene, often
used in medical imaging for comparing images taken at different times
or with different modalities.
8. Panorama Stitching: Combining multiple images to create a single,
wide-angle or panoramic view, used in photography and virtual tours.
9. Biometric Identification: Image processing is used in biometric systems
for fingerprint recognition, iris scanning, and hand geometry.
10.Remote Sensing: Analyzing satellite and aerial imagery for applications
in agriculture, urban planning, and environmental monitoring.
 18

IMAGE PROCESSING EXAMPLE

Image Denoising:
• Description: Image denoising is the process of removing unwanted noise or artifacts from an image,
resulting in a cleaner and more visually appealing representation of the original content. Noise can be
caused by factors such as low lighting conditions, sensor limitations in cameras, or transmission errors
in digital images.
• Image Processing Techniques:
1. Filtering: Filtering techniques such as median filtering or Gaussian smoothing are commonly used to
reduce noise. Median filtering replaces each pixel value with the median value of its local neighborhood,
which can effectively remove impulsive noise. Gaussian smoothing applies a weighted average to each
pixel's neighborhood, reducing high-frequency noise.
2. Wavelet Transform: The wavelet transform is used for multiresolution analysis, and it helps to separate
image content from noise in different scales and orientations. This makes it useful for denoising
applications.
3. Non-local Means Denoising: This technique compares similar patches of the image to estimate the noise
level and then reduces noise while preserving image details. It is effective in reducing random noise.
 19

IMAGE PROCESSING EXAMPLE

• Example:
Imagine you have a digital photograph taken in low light conditions. As a result,
the image contains noticeable speckles, which are a type of noise. To improve the
image, you can apply image denoising techniques:
1. Gaussian Smoothing: You apply a Gaussian filter to the image, which blurs it
slightly, reducing high-frequency noise. The resulting image appears
smoother and cleaner.
2. Wavelet Transform: Using a wavelet denoising algorithm, you decompose the
image into different scales and remove noise in the high-frequency
components while preserving the main features.
3. Non-local Means Denoising: This technique identifies similar patches in the
https://www.ques10.com/p/33595/what-is-image-processing-explain-fundamental-steps/
20

FUNDAMENTALS STEPS IN IMAGE PROCESSING

https://medium.com/futframe-ai/fundamental-steps-of-digital-image-processing-d7518d6bb23c
https://www.simplilearn.com/image-processing-article
1. Image Acquisition:
1. Image acquisition is the process of capturing an image using devices like cameras,
scanners, or sensors.
2. It's the first step where the raw data (pixels) is obtained from the real world or a
digital source.
2. Preprocessing:
Image preprocessing involves operations to prepare the image for subsequent
analysis or enhancement.
Common preprocessing steps include:
• Image Scaling: Resizing the image to a specific dimension.
• Image Cropping: Removing unwanted parts of the image.
• Color Conversion: Converting between color spaces (e.g., RGB to grayscale).
• Noise Reduction: Reducing noise using filters or techniques like averaging.
• Image Enhancement: Adjusting contrast, brightness, and sharpness.
 21

FUNDAMENTALS STEPS IN IMAGE PROCESSING

https://developer.ibm.com/articles/learn-the-basics-of-computer-vision-and-object-detection/

3. Image Segmentation:
• Image segmentation divides the image into meaningful regions or objects.
• It helps identify and isolate areas of interest within the image.
• Techniques like thresholding, edge detection, or clustering are used for
segmentation.
4. Feature Extraction:
• Feature extraction involves selecting relevant information or attributes
from the segmented regions.
• These features can be characteristics like color, texture, shape, or edges.
• Feature extraction is crucial for subsequent analysis and classification
tasks.
 22

FUNDAMENTALS STEPS IN IMAGE PROCESSING

5. Image Analysis:
• Image analysis encompasses tasks like object recognition, classification, or
measurement.
• This step uses the extracted features to derive information or make decisions.
• Machine learning, pattern recognition, and computer vision techniques are often
employed here.
6. Post-processing:
• Post-processing may involve further adjustments or operations to refine the results.
• Techniques like filtering, morphological operations, and spatial transformations can be
applied.
7. Visualization and Display:
• The final processed image can be displayed, saved, or transmitted for interpretation and
decision-making.
• Visualization can include generating visual representations, reports, or interactive
displays.
 23

FUNDAMENTALS STEPS IN IMAGE PROCESSING

8. Storage and Management:

• Processed images are often stored in databases or file systems for future
reference and retrieval.
• Effective management ensures that images are organized and accessible.
9. Interpretation and Decision-Making:
• The ultimate goal of image processing is often to extract information,
make decisions, or gain insights from the processed images.
• Interpretation can be performed by humans or automated systems.
 24

EXAMPLE
 OBJECT RECOGNITION
25

Object recognition is the technique of identifying the object present in images and videos.
 26

DETECTION
In Image classification, This algorithm locates Object Detection algorithms
it takes an image as an the presence of an act as a combination of
input and outputs the object in the image and image classification and
classification label of represents it with a object localization.
that image bounding box.
 27

IMAGE SEGMENTATION
 28

IMAGE SEGMENTATION
 29

IMAGE SEGMENTATION
 30

IMAGE SEGMENTATION
 31

https://www.youtube.com/watch?v=5QUmlXBb0MY

IMAGE SEGMENTATION
 32

OCR
 33

ROBO
MEDITCAL IMAGE ANALYSIS
34
CV IN AGRICULTURE
35
CV IN INDUSTRIAL QUALITY 36

CONTROL
IMAGE ENHANCEMENT
37
 38

IMAGE COMPRESSION
 39

IMAGE RESTORATION
 40

IMAGE FILTERING
 41

COLOR CORRECTION
 42

MORPHOLOGICAL OPERATIONS
 IMAGE REGISTRATION
43

Image registration is the process of transforming different sets of data into one coordinate
system. Data may be multiple photographs, data from different sensors, times, depths, or
viewpoints.
 44

PANORAMA STITCHING
 45

BIOMETRIC IDENTIFICATION
 46

REMOTE SENSING
THANK YOU
Dr.S.Nagarajan