UNIT–5: COMPUTER VISION
1. COMPUTER VISION VS IMAGE PROCESSING
2. APPLICATIONS OF COMPUTER VISION –
i. AGRICULTURAL MONITORING
ii. SURVEILLANCE CAMERAS
iii. DRIVERLESS CARS / SELF -DRIVER VEHICLES
iv. GOOGLE LENS
v. IMAGE BASED SEARCH ON GOOGLE
vi. FACIAL RECOGNITION
vii. FACE FILTERS
viii. Computer Vision in Retail
ix. GOOGLE TRANSLATE APP
3. COMPUTER VISION TASKS
a. FOR SINGLE OBJECTS
i. CLASSIFICATION
ii. CLASSIFICATION+LOCALISATION
b. FOR MULTIPLE OBJECTS
i. OBJECT DETECTION
ii. INSTANCE SEGMENTATION
4. BASICS OF IMAGE
a. PIXELS
b. Resolution
c. Pixel value
5. TYPES OF IMAGES –
1. GRAYSCALE IMAGES
2. RGB IMAGES
MIND MAP OF THE UNIT
VOCABULARY OF THE UNIT
1. COMPUTER VISION VS IMAGE PROCESSING
2. CLASSIFICATION
3. CLASSIFICATION+LOCALISATION
4. OBJECT DETECTION
5. INSTANCE SEGMENTATION
6. PIXELS
7. Resolution
8. Pixel value
9. GRAYSCALE IMAGES
10. RGB IMAGES
ONE-SENTENCE DEFINITIONS FOR EACH TERM:
1. Computer Vision vs Image Processing – Image Processing enhances or modifies
images, while Computer Vision enables machines to understand and interpret
visual information from images or videos.
2. Classification – Classification is the task of assigning an entire image to a single
category or label.
3. Classification + Localisation – It identifies the object in an image and also draws
a bounding box around its location.
4. Object Detection – Object Detection identifies and locates multiple objects in an
image using bounding boxes and labels.
5. Instance Segmentation – Instance Segmentation detects objects and outlines
each object separately at the pixel level.
6. Pixels – Pixels are the smallest individual units of a digital image that represent
color or intensity.
7. Resolution – Resolution refers to the number of pixels in an image, determining
its clarity and detail.
8. Pixel Value – Pixel value represents the intensity or color information stored in a
pixel.
9. Grayscale Images – Grayscale images contain only shades of black, white, and
grey without any color.
[Link] Images – RGB images use three color channels—Red, Green, and Blue—to
produce colored images.
COMPUTER VISION – MIND MAP KEYWORDS
IMAGE & COMPUTER VISION CONCEPTS
Computer Vision
→ Understanding images
→ Object recognition
→ Decision making
Image Processing
→ Image enhancement
→ Filtering
→ Noise removal
Classification
→ Single label
→ Whole image
→ Category prediction
Classification + Localisation
→ Single object
→ Label + bounding box
→ Object position
Object Detection
→ Multiple objects
→ Bounding boxes
→ Labels
Instance Segmentation
→ Pixel-level detection
→ Separate object masks
→ Precise boundaries
Pixels
→ Smallest unit
→ Image building block
→ Color dot
Resolution
→ Total pixels
→ Width × Height
→ Image clarity
Pixel Value
→ Intensity level
→ Color information
→ 0–255 range
Grayscale Images
→ Black & white
→ Single channel
→ Shades of grey
RGB Images
→ Red channel
→ Green channel
→ Blue channel
→ 3 channels