Top Banner
Human Vision Nov 15.2012 Wu Pei
27

Human Vision

Feb 24, 2016

Download

Documents

trella

Human Vision. Nov 15.2012 Wu Pei. The eye. Optical system. Cone and Rod Vision:. In a sense, each of us has two visual systems; the photopic system functions in lighted conditions; the scoptopic system functions in dim light - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Human Vision

Human Vision

Nov 15.2012Wu Pei

Page 2: Human Vision

The eye

Page 3: Human Vision

Optical system

Page 4: Human Vision
Page 5: Human Vision

Cone and Rod Vision:

• In a sense, each of us has two visual systems; the photopic system functions in lighted conditions; the scoptopic system functions in dim light

• Cones mediate photopic vision; rods mediate scotopic vision; this is called the duplexity theory of vision

Page 6: Human Vision

Distribution of cone and rod

Page 7: Human Vision

Blind point

Page 8: Human Vision

Spatial resolution

• 4.5 million cone cells and 90 million rod cells in the human retina in average.

• High resolution in middle and low resolution in edge, highest at fovea, about 1.2'.

• Field of view: best resolution horizontal 35°and vertical 20°

• In all: 1750*1000 in best resolution zone. About 5-15 megapixel whole field

Page 9: Human Vision

Qualitative representation of visual detail using a single glance of the eyes.

Page 10: Human Vision

Time resolution

• Persistence of vision: an afterimage is thought to persist for approximately one twenty-fifth of a second on retina.

• Fps: frames per second• The mechanism of film system.

Page 11: Human Vision

Why game need better fps than movie

• Movie: 24fps TV: 25fps• Game: 59/60fps

Page 12: Human Vision

Answer

Real video Game

Page 13: Human Vision

Sensitivity to light• Human eye can is self-adaptive to differed

light intensity by adjusting rhodopsin concentration.

Page 14: Human Vision

Light sensitivity

Page 15: Human Vision

Evolutionary perspective

• Our vision is optimized for receiving the most abundant spectral radiance our star emits.

Page 16: Human Vision

Compared with camera

Page 17: Human Vision

Contrast sensitivity

Page 18: Human Vision

Color Depth

• True color :8*3=24 bits• HDR: 32*3=96 bits• HDR use float but not fixed point number for

luminance ranges from 10-4 to 108 . In practice 10bit to encode luminance and 8bit for hue is enough for eliminating quantization error.

Page 19: Human Vision

Generating HDR Image

Page 20: Human Vision

Color banding

Page 21: Human Vision

Hints for video coding

• Human eye is less sensitive to chrominance signal than to luminance signal (U and V can be coarsely coded)

• Human eye is less sensitive to the higher spatial frequency components

• Human eye is less sensitive to quantizing distortion at high luminance levels

Page 22: Human Vision

Estimating bandwidth of eye

• Input:24bit*10Mpixel*24fps=720MB/s• Output: ~100kB/s• Compare: H.264,1080p:1.5MB/s

Page 23: Human Vision

How the visual information encoded

• The process of neuron development is a process of “training”.

Page 24: Human Vision

Encoding and illusion

• Illusion means visually perceived images that differ from objective reality.

• Generally visual encoding is a lossy coding. So look at the illusion will be helpful to know the encoding process.

Page 25: Human Vision

Mach band: Lateral inhibition

Page 26: Human Vision

More example: Hermann grid

Page 27: Human Vision

Thank You