Scribd
Upload
16 views17 pages

Mid Note 2

Uploaded by

weird Al
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
16 views17 pages

Mid Note 2

Uploaded by

weird Al
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 17

Lecture 2 :

Properties of light -->What is light?


Electromagnetic radiation (EMR) moving along rays in space
Plenoptic Function
• We can describe all of the light in a scene by specifying the radiation from all points in
space and from all directions.

The light field


- Known as the plenoptic function
The visible light spectrum
- We see electromagnetic radiation in a small range of wavelengths.
Light response is nonlinear
Our visual system has a large dynamic range
• We can resolve both light and dark things at the same time
• One mechanism for achieving this is that we sense light intensity on a
logarithmic scale
– an exponential intensity ramp will be seen as a linear ramp
• Another mechanism is adaptation
– rods and cones adapt to be more sensitive in low light, less sensitive in
bright light.

Rods and cones act as filters on the spectrum


(SPD)
• To get the output of a filter, multiply its
response curve by the spectrum,
integrate over all wavelengths
– Each cone yields one number
• Q: How can we represent an entire
spectrum with 3 numbers?
sRGB
● industry standard encoding of colors for most cameras, printers, displays, etc. Most images are encoded
as sRGB.

Adding a lens:
A lens focuses light onto the film
- There is a specific distance at which objects are “in focus” (other points project to a “circle of
confusion” in the image)
- Changing the shape of the lens changes this distance

Lenses
A lens focuses parallel rays onto a single focal point at a distance f, with an aperture of D limiting the
range of rays.
Bayer filters
• Bayer filter is a color filter array for arranging RGB filters on a
photosensor grid
• It typically has a pattern of 50% green, 25% red, 25% blue
• The green → higher resolution B&W image
• The red and blue →create a lower-resolution color image
• The process of reconstructing a full-color image from these
incomplete samples is called demosaicking or interpolation.
Lecture3
Lecture4

- Interpolation is the process of increasing the size of an image by adding new pixels and estimating
their values based on the values of surrounding pixels. The goal is to produce a larger image with
higher resolution without losing details or introducing noise.

- Decimation is a technique used to reduce the computational complexity of image processing


algorithms and reduce the storage requirements of large images

General resizing:-
◼ This technique involves changing the size of an image by stretching or shrinking the
image to the desired dimensions.
◼ It can be done by keeping the aspect ratio of the image unchanged or changing it as per
the requirement.
.
z

Multi-resolution image pyramids


• Commonly used in coarse-to-fine
matching, optical flow,
stereo, blending, …
• … deep neural networks

Image blending pyramids


Multi-scale pyramids blend two or more images seamlessly, from low-resolution to high-resolution, using
a mask to control the degree of blending. This can be used for image compositing, panoramic stitching,
and video stabilization.
Image resampling
• Changes size or resolution of an image
• Interpolation methods used: nearest neighbor, bilinear, bicubic, spline
• Choice s epends on quality and computational requirements
• Used for: upsampling, downsampling, changing aspect ratio
• Fundamental operation in image processing
• Used in computer vision, digital photography, and video processing.

• Transformation of an image's geometry

• Changes relationship between pixels

• used for perspective correction, face recognition,


object tracking, etc.

• Requires knowledge of the mapping between the


input and output image
Data fitting and optimization
Demosaicing:
Demosaicing, also known as debayering, is a process in image processing that converts raw image data
captured by color sensors into a full-color image. Most color sensors in digital cameras use a mosaic filter
pattern to capture only one color component for each pixel, but the missing color information for each
pixel must be reconstructed from the surrounding pixels using various techniques. The quality of the
demosaiced image depends on the accuracy of the algorithm used and can be affected by factors such as
noise, the type of mosaic filter used, and the subject matter of the image.
Lecture5
Hyperparameters:-
What is the best value of K to use? What is the best distance metric to use? These are examples of hyperparameters:
choices about our learning algorithm that we don’t learn from the training data; instead we set them at the start of the
learning process Very problem-dependent. In general need to try them all and see what works best for our data / task. I

You might also like