FlatCV

Morphological Operations

Binary morphological operations for processing black-and-white images. These operations use a disk-shaped structuring element.

Erode

Erosion shrinks white regions by removing pixels at boundaries. Useful for removing small white noise or separating connected objects.

Input	Output

$ ./flatcv imgs/parrot_threshold.png erode 2 imgs/parrot_threshold_erode.png
Loaded image: 512x384 with 4 channels
Executing pipeline with 1 operations:
Applying operation: erode with parameter: 2.00
  → Completed in \d+.\d+ ms \(output: 512x384\) (regex)
Final output dimensions: 512x384
Successfully saved processed image to 'imgs/parrot_threshold_erode.png'

$ git diff --quiet imgs/parrot_threshold_erode.png

Dilate

Dilation expands white regions by adding pixels at boundaries. Useful for filling small holes or connecting nearby objects.

Input	Output

$ ./flatcv imgs/parrot_threshold.png dilate 2 imgs/parrot_threshold_dilate.png
Loaded image: 512x384 with 4 channels
Executing pipeline with 1 operations:
Applying operation: dilate with parameter: 2.00
  → Completed in \d+.\d+ ms \(output: 512x384\) (regex)
Final output dimensions: 512x384
Successfully saved processed image to 'imgs/parrot_threshold_dilate.png'

$ git diff --quiet imgs/parrot_threshold_dilate.png

Close

Closing performs dilation followed by erosion. It closes small dark gaps and holes within white regions while preserving overall shape.

Input	Output

$ ./flatcv imgs/parrot_threshold.png close 2 imgs/parrot_threshold_close.png
Loaded image: 512x384 with 4 channels
Executing pipeline with 1 operations:
Applying operation: close with parameter: 2.00
  → Completed in \d+.\d+ ms \(output: 512x384\) (regex)
Final output dimensions: 512x384
Successfully saved processed image to 'imgs/parrot_threshold_close.png'

$ git diff --quiet imgs/parrot_threshold_close.png

Open

Opening performs erosion followed by dilation. It removes small white noise and thin protrusions while preserving overall shape.

Input	Output

$ ./flatcv imgs/parrot_threshold.png open 2 imgs/parrot_threshold_open.png
Loaded image: 512x384 with 4 channels
Executing pipeline with 1 operations:
Applying operation: open with parameter: 2.00
  → Completed in \d+.\d+ ms \(output: 512x384\) (regex)
Final output dimensions: 512x384
Successfully saved processed image to 'imgs/parrot_threshold_open.png'

$ git diff --quiet imgs/parrot_threshold_open.png