Week 10: 30/11/2012

Lecture:  

Note:  This lecture has not been made where the text and images are merged into one image.  This lecture has been made separately, so I can type more for this one.  

The human eye is a very sensitive organ and can play tricks on its host.  Many famous pieces of art have been drawn/painted and when looked upon, look as though they are moving.  For example:  (See figure 1)

Figure 1 - Still Image - Does It Look As Though It Is Moving To You?  

It was believed that when the human eye see's an image, it holds that image (as if imprinted) for a 25th of a second, and if another image was produced in that time or less from the first image being shown, it would look like a moving image.  This is called the persistence of vision, which is an old idea.  It is now called the myth of the persistence of vision.
"A more plausible theory to explain motion perception (at least on a descriptive level) are two distinct perceptual illusions: phi phenomenon and beta movement."  -  From PowerPoint lecture  

As said, if one image is produced after another in less than a 25th of a second, this gives the illusion of movement.  (See figure 2)  

Figure 2 - Frames That Are 0.04 Seconds Apart  

If a video file was to be saved in an uncompressed format, this could mean you would need one of the biggest storage spaces (if not the biggest) in the world.  




—"Uncompressed  HD Video files could be large. say 3bytes per pixel, 1920x1080 by 60 frames per second = 373.2 Mbytes per second.
—i.e. approximately 1Gbyte every 3 seconds."  -  From PowerPoint lecture  
Even to today's standards, that is far too much storage.  So this is why many compression algorithms and standards are able to be used/to be put in place and reduced greatly in size.  

The term bit rate refers to the number of bits per second that are used to represent a video file (or at least any portion of a file that is video).  The range of bit rate can go from 300Kb per second to 8, 000Kb per second.  The higher the bit rate, the better the quality of a video.  
"Interlaced video is a way to make the best use of the limited bandwidth for video transmission, especially in the old days of analogue broadcasts. The receiver (your TV) "tricks" your eyes by drawing first the odd number lines on the screen 25 times per second. Then the even lines of the next frame and so on. Progressive video does not interlace and appears sharper."  -  From PowerPoint lecture

Resolution refers to the number of pixels that can be represented on the display device.  The higher the resolution, the more pixels that can be represented.  

There are many different video formats.  A website that shows a list of them and a description of them is:  http://www.libtiff.org/video-formats.html

The more you compress a file, the more data the files losses.  Due to compression artifacts, images become more distored.
For algorithms that are made to compress video predictively can still have issues with fast-paced, unpredictable, detailed motion (like sport).
The solution to this could be automatic video quality assessment.

As said before, image degradation can occur due to too much compression.  But it can also happen because of other factors.  (See figure's 3 and 4)

Figure 3 - Distorted Image Due to Lossy Compression - Compressed Image  

Figure 4 - Distorted Image Due to Camera Lens Blurring - Blurred Image  

Is it possible for a computer to decide if an image is of good quality?  (See figure 5)

Figure 5 - Different Qualities of the Same Image, With Their Quality Values?  

The thing that holds the key - statistical algorithmic video processing.


Lab:  

In the lab, my lecturer asked the class to look at the tutorial video's on the Adobe website for their product Adobe Premier Pro CS4.  This was in preparation for the video we are to make using clips and audio given to us.