Home / Rigel SDK / X-ray Image Processing & Analysis

MODULE 02 — IMAGE PROCESSING & ANALYSIS

X-ray Image Processing & Analysis

From X-ray-specific enhancement filters that define diagnostic and inspection image quality
to noise reduction, detector correction, and quantitative image quality analysis.

X-ray Enhancement

Enhancement technology developed in-house to handle the wide dynamic range of X-ray images.

NADE

Non-linear Adaptive Detail Enhancement. Multi-layer enhancement with simultaneous control of detail contrast, sharpness, noise suppression, and halo reduction.

Noise Reduction

Proven filters that suppress noise while preserving edges — including filters dedicated to battery inspection lines.

BM3D

High-quality noise reduction based on block-matching 3D collaborative filtering.

NLM

Non-local means noise reduction that averages similar neighborhoods found across the whole image.

TV Denoising

Total-variation-based noise reduction that strongly smooths flat regions while keeping boundaries.

Guided · Bilateral · Anisotropic Diffusion

Three lightweight edge-preserving filters that run in milliseconds.

AxiClearBattery-specific

Dedicated to cylindrical (jelly-roll) battery cross-sections. Noise reduction that analyzes electrode periodicity and contrast to set its strength automatically.

VertClearBattery-specific

Dedicated to prismatic and pouch batteries. Recognizes vertical electrode structures and adjusts strength by region.

Denoiser ComparisonComparison guide

Seven noise reduction filters compared under identical measured conditions — a guide to what to use when.

Detector Correction

Corrects detector-induced distortions right at acquisition. Quantitative evaluation of correction quality (FFC · BPC) is provided under Image Quality Analysis.

Flat Field Correction

Linear (single-gain) and polynomial (multi-gain) gain correction.

Bad Pixel Correction

Automatic bad pixel detection and correction.

Image Quality Analysis

Standard image quality analyzers for system development and quality control.

MTF

Modulation transfer function (MTF) analysis. Quantifies, by frequency, how sharply the system renders fine structure.

NPS

Noise power spectrum (NPS) analysis. Quantifies the amount and texture of noise by frequency.

DQE

Detective quantum efficiency (DQE) analysis. Combines MTF and NPS measurements into a single dose-versus-image-quality curve.

SRb

Automatically measures basic spatial resolution (SRb) from duplex wire gauge images, based on automatic wire pair detection and dip judgment.

SNR · CNR · gCNR

ROI-based SNR and dSNRn (normalized SNR) analysis, plus CNR and gCNR (distribution overlap).

FFC

Evaluates gain correction quality from flat images — quantifies SNR, non-uniformity, and exposure linearity.

BPC

Detects and classifies bad pixels by type (dead, over/under-responding, non-uniform, noisy, lag) and automatically generates a diagnostic report with defect statistics, judgments, and recommendations.

General Processing Tools

Basic building blocks for image processing algorithms. Inspection blocks such as segmentation and matching are covered in the 2D/3D Inspection module.

Basic Operations · Image Objects

Multi-dimensional image containers and type conversion, crop/pad/resize/binning, arithmetic, logic, and mask operations, statistics and profile extraction — the basic elements of a pipeline.

Geometric Transforms

Coordinate transform tools including flip and rotation, affine and perspective warps, polar and log-polar transforms, B-spline deformation fields, and reslicing.

Enhancement · Histogram

Contrast stretching, gamma and sigmoid remapping, unsharp masking, local contrast enhancement, and histogram equalization, CLAHE, and matching.

Spatial Filters · Morphology

Arbitrary and separable convolution with kernel design, Gaussian/median/moving-average filters, DoG and LoG, edge detection such as Canny and Sobel with gradients and structure tensors, and morphological operations such as erosion and dilation.

Frequency Processing

1D/2D/3D FFT with window functions, Gaussian and Butterworth frequency filters, frequency-domain unsharp masking, and inverse deconvolution.

Miscellaneous

Auxiliary tools outside the main categories — thinning, focus (sharpness) measurement, PVE correction, and point-set sampling.

Need X-ray Image Processing & Analysis technology?

We will guide you to the right configuration and an evaluation SDK for your system or software.

Contact us