Infographics / Digital Radiography

Image Processing: Histogram, LUT & Spatial-Frequency Filtering

Left: an intensity histogram plots frequency of pixels on the y-axis against pixel value on the x-axis. Right: a look-up table transfer curve remaps input pixel value to output display grey level, with a steep central window setting contrast (window and level). Below: three edge profiles show that a low-pass filter smooths and blurs an edge while a high-pass filter (unsharp masking) sharpens it by boosting high spatial frequencies, also amplifying noise.

Digital image post-processing: histogram, look-up table (LUT), and spatial-frequency filtering Left: an intensity histogram plots frequency of pixels on the y-axis against pixel value on the x-axis. Right: a look-up table transfer curve remaps input pixel value to output display grey level, with a steep central window setting contrast (window and level). Below: three edge profiles show that a low-pass filter smooths and blurs an edge while a high-pass filter (unsharp masking) sharpens it by boosting high spatial frequencies, also amplifying noise. Grey-scale mapping: histogram feeds a look-up table (LUT) Histogram analysis pixel value (grey level) frequency (no. of pixels) Look-up table (transfer curve) window (contrast) level LUT input pixel value output grey level Spatial-frequency filtering (convolution kernels) on an edge input: sharp edge position across edge low-pass: smooth / blur high frequencies attenuated high-pass: sharpen / edge over/undershoot boosts edge High-pass / unsharp masking boosts high spatial frequencies: edges sharpen but noise also increases. LUT / window-level changes display contrast only — it does not alter spatial resolution (pixel size, MTF).
Practise Digital RadiographyDownload SVG

More in Digital Radiography

Practice the Physics paper

FRCRX is a Part 1 Physics question bank: true/false items with explanations, timed mocks and topic analytics.

Start question practiceSee plans