LabKit MATLAB Workbench

app

Image Enhance

Image Enhance builds an ordered, reversible processing history for one image or a batch and exports the resulting images with the exact step sequence.

Requirements And Launch

All processing uses MATLAB and repository-owned code.

labkit_ImageEnhance_app

Inputs And Batch Mode

Add supported image files or a folder. In Batch shared processing mode, one shared step history is applied to every source. When batch mode is off, each image keeps a separate history and optional white ROI. Selecting another image updates the preview without recalculating unrelated files.

Basic Workflow

  1. Load images and select a representative source.
  2. Decide whether steps are shared across the batch.
  3. Choose a tool and set its controls.
  4. For White ROI calibration, draw the white-background ROI first.
  5. Apply the tool to history and inspect Enhanced, Original, or Before | After.
  6. Undo or reset history as needed.
  7. Choose output format/folder and export.

Each Apply tool action saves one processing step. Panner changes preview the pending tool without changing the saved history until Apply tool is chosen. ROI creation and overlay refresh preserve the current zoom.

Tools And Defaults

ToolPrimary controlSecondary control
Brightness/contrastBrightness 0%Contrast 15%
Local contrastClarity 30%Radius 12 px
SharpenSharpen 35%Radius 1.5 px
Hue/saturationHue 0 degreesSaturation 10%
White balanceStrength 100%Temperature 0%
White ROI calibrationStrength 100%White target 92%
Subject-preserving enhanceStrength 70%Background target 90%

Processing converts input to RGB double in [0,1], applies steps in table order, and clamps the final image. Brightness, local contrast, sharpening, hue/saturation, and gray-world white balance call the linked labkit.image primitives.

White ROI calibration requires a valid rectangle for every affected image. It builds a softened background mask from the rectangle and adjusts luminance toward the target while limiting subject changes. Subject-preserving enhance estimates a low-saturation bright background automatically and uses a whole image fallback only when no usable background support is found.

History And Reproducibility

The history table records step kind and settings in execution order. Undo history removes the latest applied step; Reset history returns to the normalized original. Source images are not modified. Shared and per-image histories are stored separately in project state so switching batch mode does not ambiguously merge them.

Outputs

Export supports PNG, TIFF, and JPEG. Each output is rendered from the source plus its effective stored pipeline, not from a downsampled preview. The export manifest records source, ordered steps, ROI geometry, batch mode, format, and output filename. A LabKit result JSON records the complete output set and the project parameters used to create it.

Use Without The GUI

steps = [ ...
    image_enhance.analysisRun.makeStep("Brightness/contrast", 0, 15), ...
    image_enhance.analysisRun.makeStep("Local contrast", 30, 12), ...
    image_enhance.analysisRun.makeStep("Sharpen", 35, 1.5)];
output = image_enhance.analysisRun.applyPipeline(imread("source.png"), steps);
imwrite(output{1}, "enhanced.png");

Project And State

Saved projects keep portable source references, shared and per-image step histories, white-reference ROIs, export settings, and compact result metadata. Full-size pixels and downsampled previews are loaded again from the selected source when a project opens rather than embedded in the project.

An empty launch does not choose an output directory. Adding images establishes the source-adjacent default; Choose folder remains available before export.

Errors And Limitations

Functions And API

Open a function for exact MATLAB syntax, arguments, outputs, behavior, and source.

FunctionPurpose
image_enhance.analysisRun.applyPipelineReplay an ordered non-destructive image enhancement history.

Change history