Several conceptual image Digitization 1 sampling 2 quantization 3 Digital Image Properties 31 Digital Image measurement and topological sequence 2 histogram 3 entropy 4 Image visual perception 5 image quality 6 image noise color image 1 Color Physics 2 color 3 colour space
1 Some conceptsImage: We usually intuitively understand its meaning, such as images on the retina of the human eye or images captured by a TV camera. Brightness (brightness): the integration of different optical quantities, and the use of luminance as a basic amount allows us to avoid describing the imaging process of the image. Luminance images (intensity image): This is a 2D image that records the luminance information. Perspective projection (Perspective projection): U=xfz u = x f z u = \frac{xf}{z} V=yfz v = y F z v = \frac{yf}{z} grayscale (gray-level): Lowest value corresponds in monochrome image To black, while the highest value corresponds to white. The grayscale value between them is gray scale. The quality of digital image increases with the increase of space, spectrum, radiation metering and time resolution
Spatial resolution (spatial resolution): determined by the proximity between image sampling points on the image plane. Spectral resolution (spectral resolution): determined by the sensor's light frequency bandwidth. Radiometric Measurement Resolution (radiometric resolution): Corresponds to the number of distinguishable gray levels. Time resolution: Depends on the image acquisition time sampling interval.2 Image DigitizationSampling (sampled): The Matrix of f (x, Y) f (y), F (X.) is sampled as one MXN MXN m\times N. Image quantization (quantization): Gives an integer number to each successive sample value.2.1 SamplingA continuous image is digitized at the sampling point, which is arranged on a plane, and its relations are calledGrid (GRID). Raster (Raster): An adjacent raster is defined between points. Pixel (pixel) (image Element): An infinitely small sample point in the grid corresponds to a cell in the digitized image.2.2 QuantificationQuantization (quantization): Sample image Value FS (jδx,kδy) F S (Jδx, Kδy) f_s (J\delta X,k\delta y) is represented by a number. The process of transforming the continuous numerical value (brightness) of an image function into its numerical equivalent. Quantization: The B-bit represents the value of the image brightness, then the brightness order is k=2b k = 2 b k = 2^b.2.3 Digital Image Properties 2.3.1 Measurement and topological sequence of digital imagesConditions for distance satisfaction
Identity: D (p,q) ≥0, when and only if P=q D (p,q) =0 D (p, Q) ≥0, when and only if p = q is D (p, q) = 0 d (p,q) \ge 0, when and only if \ P =q \quad D (p,q ) = 0 Symmetry: D (p,q) =d (q,p) d (p, q) = d (q, p) d (p,q) =d (q,p) Triangular inequalities: D (P,