Source Steven Seitz Convention larger values correspond to brighter content A

Source steven seitz convention larger values

This preview shows page 100 - 112 out of 124 pages.

(Source: Steven Seitz) Convention: larger values correspond to brighter content. A color image is defined similarly as a 3 component vector-valued function: x ( s 1 , s 2 ) = r ( s 1 , s 2 ) g ( s 1 , s 2 ) b ( s 1 , s 2 ) . 78
Image of page 100

Subscribe to view the full document.

Image representation – Types of images – Digital imagery Digital imagery Raster images Sampling: reduce the 2d continuous space to a discrete grid Ω Z 2 Gray level image: Ω R (discrete position to gray level) Color image: Ω R 3 (discrete position to RGB) 79
Image of page 101
Image representation – Types of images – Digital imagery Bitmap image Quantization: map each value to a discrete set [0 , L - 1] of L values ( e.g., round to nearest integer) Often L = 2 8 = 256 ( 8 bit images unsigned char ) Gray level image: Ω [0 , 255] ( 255 = 2 8 - 1 ) Color image: Ω [0 , 255] 3 Optional: assign instead an index to each pixel pointing to a color palette (format: .png , .bmp ) 80
Image of page 102

Subscribe to view the full document.

Image representation – Types of images – Digital imagery Digital imagery Digital images: sampling + quantization: -→ 8bit images can be seen as a matrix of integer values We will refer to an element s Ω as a pixel location, x ( s ) as a pixel value, and the pair ( s, x ( s )) as a pixel (“picture element”). 81
Image of page 103
Image representation – Types of images – Digital imagery Functional representation: x : Ω Z d R K d : dimension ( d = 2 for pictures, d = 3 for videos, . . . ) K : number of channels ( K = 1 monochrome, 3 colors, . . . ) s = ( i, j ) : pixel position in Ω x ( s ) = x ( i, j ) : pixel value(s) in R K 82
Image of page 104

Subscribe to view the full document.

Image representation – Types of images – Digital imagery Functional representation: x : Ω Z d R K d : dimension ( d = 2 for pictures, d = 3 for videos, . . . ) K : number of channels ( K = 1 monochrome, 3 colors, . . . ) s = ( i, j ) : pixel position in Ω x ( s ) = x ( i, j ) : pixel value(s) in R K Array representation ( d = 2 ): x ( R K ) n 1 × n 2 n 1 × n 2 : n 1 : image height, and n 2 : width x i,j R K : pixel value(s) at position s = ( i, j ) : x i,j = x ( i, j ) 82
Image of page 105
Image representation – Types of images – Digital imagery For d > 2 , we speak of multidimensional arrays : x ( R K ) n 1 × ... × n d d is called dimension, rank or order, In the deep learning community: they are referred to as tensors (not to be confused with tensor fields or tensor imagery). 83
Image of page 106

Subscribe to view the full document.

Image representation – Types of images – Digital imagery Vector representation: x ( R K ) n n = n 1 × n 2 : image size (number of pixels) x k R K : value(s) of the k -th pixel at position s k : x k = x ( s k ) 84
Image of page 107
Image representation – Types of images – Digital imagery Color 2d image: Ω Z 2 [0 , 255] 3 Red, Green, Blue (RGB), K = 3 RGB: Usual colorspace for acquisition and display There exist other colorspaces for different purposes: HSV (Hue, Saturation, Value), YUV, YCbCr. . . 85
Image of page 108

Subscribe to view the full document.

Image representation – Types of images – Digital imagery Spectral image: Ω Z 2 R K Each of the K channels is a wavelength band For K 10 : multi-spectral imagery For K 200 : hyper-spectral imagery Used in astronomy, surveillance, mineralogy, agriculture, chemistry 86
Image of page 109
Image representation – Types of images – Digital imagery The Horse in Motion (1878, Eadweard Muybridge) Gray level video: Ω Z 3 R 2 dimensions for space 1 dimension for time 87
Image of page 110

Subscribe to view the full document.

Image representation – Types of images – Digital imagery
Image of page 111
Image of page 112

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern

Ask Expert Tutors You can ask 0 bonus questions You can ask 0 questions (0 expire soon) You can ask 0 questions (will expire )
Answers in as fast as 15 minutes