EE 4780 Image Enhancement
Bahadir K. Gunturk2 Image Enhancement The objective of image enhancement is to process an image so that the result is more suitable than the original image for a specific application. There are two main approaches: Image enhancement in spatial domain: Direct manipulation of pixels in an image Point processing: Change pixel intensities Spatial filtering Image enhancement in frequency domain: Modifying the Fourier transform of an image
Bahadir K. Gunturk3 Image Enhancement by Point Processing Intensity Transformation
Bahadir K. Gunturk4 Image Enhancement by Point Processing Contrast Stretching
Bahadir K. Gunturk5 Image Enhancement by Point Processing Contrast Stretching
Bahadir K. Gunturk6 Image Enhancement by Point Processing Intensity Transformation Matlab exercise
Bahadir K. Gunturk7 Image Enhancement by Point Processing Intensity Transformation
Bahadir K. Gunturk8 Image Enhancement by Point Processing Intensity Transformation
Bahadir K. Gunturk9 Image Enhancement by Point Processing Gray-Level Slicing
Bahadir K. Gunturk10 Image Enhancement by Point Processing Histogram 0 255
Bahadir K. Gunturk11 Histogram Specification Intensity mapping Assume T(r) is single-valued and monotonically increasing. The original and transformed intensities can be characterized by their probability density functions (PDFs)
Bahadir K. Gunturk12 Histogram Specification The relationship between the PDFs is Consider the mapping Cumulative distribution function of r Histogram equalization!
Bahadir K. Gunturk13 Image Enhancement by Point Processing Histogram Equalization
Bahadir K. Gunturk14 Image Enhancement by Point Processing Histogram Equalization Example Intensity Number of pixels Intensity Number of pixels
Bahadir K. Gunturk15 Image Enhancement by Point Processing Histogram Equalization
Bahadir K. Gunturk16 Histogram Specification
Bahadir K. Gunturk17 Histogram Specification
Bahadir K. Gunturk18 Histogram Specification
Bahadir K. Gunturk19 Histogram Specification
Bahadir K. Gunturk20 Local Histogram Processing Histogram processing can be applied locally.
Bahadir K. Gunturk21 Image Subtraction The background is subtracted out, the arteries appear bright.
Bahadir K. Gunturk22 Image Averaging Original image Noise Corrupted image Assume n(x,y) a white noise with mean=0, and variance If we have a set of noisy images The noise variance in the average image is
Bahadir K. Gunturk23 Image Averaging
Bahadir K. Gunturk24 Spatial Filtering A low-pass filter A high-pass filter
Bahadir K. Gunturk25 Spatial Filtering Median Filter Sort: ( ) Example Original signal: Noisy signal: Filter by [ 1 1 1]/3: Filter by 1x3 median filter:
Bahadir K. Gunturk26 Spatial Filtering Median filters are nonlinear. Median filtering reduces noise without blurring edges and other sharp details. Median filtering is particularly effective when the noise pattern consists of strong, spike-like components. (Salt-and- pepper noise.)
Bahadir K. Gunturk27 Spatial Filtering Original 3x3 averaging filter Salt&Pepper noise added 3x3 median filter
Bahadir K. Gunturk28 Spatial Filtering
Bahadir K. Gunturk29 Wiener Filter Original image Noise Noisy image Noise variance Signal variance
Bahadir K. Gunturk30 Wiener Filter is estimated by Since variance is nonnegative, it is modified as Estimate signal variance locally: N N
Bahadir K. Gunturk31 Wiener Filter Noisy, =10 Denoised (3x3neighborhood) Mean Squared Error is 56 wiener2 in Matlab
Bahadir K. Gunturk32 Spatial Filtering Gradient Operators Averaging of pixels over a region tends to blur detail in an image. As averaging is analogous to integration, differentiation can be expected to have the opposite effect and thus sharpen an image. Gradient operators (first-order derivatives) are commonly used in image processing applications.
Bahadir K. Gunturk33 Spatial Filtering Gradient Operators These are called the Sobel operators
Bahadir K. Gunturk34 Spatial Filtering Laplacian Operators Laplacian operators are second-order derivatives.
Bahadir K. Gunturk35 Spatial Filtering
Bahadir K. Gunturk36 Spatial Filtering High-boost or high-frequency-emphasis filter Sharpens the image but does not remove the low-frequency components unlike high-pass filtering
Bahadir K. Gunturk37 Spatial Filtering High-boost or high-frequency-emphasis filter High pass = Original – Low pass High boost = (Original) + K*(High pass)
Bahadir K. Gunturk38 Spatial Filtering A high-pass filter A high-boost filter
Bahadir K. Gunturk39 Spatial Filtering High-boost or high-frequency-emphasis filter
Bahadir K. Gunturk40 Spatial Filtering