Guruh: 214-19 tayyorladi: Axrorov Asadbek

VEYVLET IXCHAMLASHTIRISH

Chapdagi rasm o'ngdagi tasvirning piksel qiymatlarini ifodalaydi, 8 × 8 tasvir. 2D Haar o'zgartirishni ikkinchi daraja uchun qo'llash tasvir hajmining chiziqli qisqarishiga olib keladi:

VEYVLET IXCHAMLASHTIRISH

Hisoblangan farqlar va tasvirning qisqarishi tasvirni sifatni saqlab qolgan holda kamroq ma'lumot bilan ixchamlash imkonini beradi.

Ko'proq ixchamlash past sifatni anglatadi va yuqori sifat kamroq ixchamlashni anglatadi.

Rangli tasvirlar uchun ham xuddi shunday qo'llaniladi:

Misol (MATLAB)

Misol (MATLAB)

1-usul. Global Thresholding.

load woman; % Load original image

image(X)

title('Original Image')

colormap(map)

n = 5; % Decomposition level

w = 'sym8'; % Near symmetric wavelet

[c,l] = wavedec2(x,n,w); % Multilevel 2-D wavelet decomposition

opt = 'gbl'; % Global threshold

thr = 20; % Threshold

sorh = 'h'; % Hard thresholding

keepapp = 1; % Approximation coefficients cannot be thresholded

[xd,cxd,lxd,perf0,perfl2] = wdencmp(opt,c,l,w,n,thr,sorh,keepapp);

image(x)

title('Original Image')

colormap(map)

figure

image(xd)

title('Compressed Image - Global Threshold = 20')

colormap(map)

2-usul. Level-Dependent Thresholding.

load woman; % Load original image

image(X)

title('Original Image')

colormap(map)

n = 5; % Decomposition level

w = 'sym8'; % Near symmetric wavelet

[c,l] = wavedec2(x,n,w); % Multilevel 2-D wavelet decomposition

opt = 'lvd'; % Level-dependent thresholds

thr_h = [17 18]; % Horizontal thresholds

thr_d = [19 20]; % Diagonal thresholds

thr_v = [21 22]; % Vertical thresholds

thr = [thr_h ; thr_d ; thr_v];

[xd2,cxd2,lxd2,perf02,perfl22] = wdencmp(opt,x,w,2,thr,sorh);

image(x)

title('Original Image')

colormap(map)

figure

image(xd2)

title('Compressed Image - Level-Dependent Thresholding')

colormap(map)

E'TIBORINGIZ UCHUN RAHMAT!