Denoising

This function block reduces noise in images using a fast non-local means denoising method. It works for both color and grayscale images and provides sliders to balance noise removal and detail preservation.

📥 Inputs

Image Any Connect an image (color or grayscale) that you want to denoise.

(hint: this is an input socket)

📤 Outputs

Image Any The denoised image is provided here.

(hint: this is an output socket)

🕹️ Controls

Strength Adjusts the overall denoising strength. Higher values remove more noise but may blur fine details.

Averaging Controls the averaging power used for smoothing. Increasing this can reduce noise but may soften edges.

Blend Noise (Colored) Adjusts color-channel blending for colored images. Use to control how color noise is treated relative to luminance noise.

(hint: these are interactive sliders visible on the block)

⚙️ Running Mechanism

• When the block receives an image through the Image Any input, it applies the selected denoising parameters and updates the output with the processed image.

• Parameters are read from the sliders each run, so changing a slider affects the next evaluation.

• The block handles both single-channel and multi-channel images, selecting the appropriate processing automatically.

✨ Features

• Works for both grayscale and color images.

• Real-time parameter tuning using sliders for instant visual feedback.

• Balanced noise removal to preserve edges when tuned correctly.

📝 Usage Instructions

1. Connect an image-producing block to the Image Any input.

2. Adjust Strength to control how aggressively noise is removed.

3. Fine-tune Averaging to manage smoothness versus detail.

4. For color images, tweak Blend Noise (Colored) to adjust color-specific denoising.

5. Observe the denoised result from the Image Any output (connect to a viewer or saver).

💡 Tips and Tricks

• For noisy camera captures, try combining this block with Bilateral Filter or Blur before denoising to reduce impulse noise and make results cleaner.

• Use Show Image to preview results live while you tune sliders.

• To save examples for later comparison, connect the output to Image Logger or Image Write.

• If you want to keep a stable frame for comparison during tweaking, use Image Memory to freeze a chosen frame while you adjust controls.

• Crop to the region of interest first with Image ROI Select to focus denoising only where it matters and save compute.

(hint: recommended complementary blocks are listed above for common workflows)

🛠️ Troubleshooting

• If the result looks overly smooth or details are lost: lower the Strength and reduce Averaging.

• If color artifacts appear after denoising: reduce Blend Noise (Colored) or try preprocessing with Blur.

• If performance is slow on large images: resize with Image Resizer or process smaller ROIs via Image ROI Select.

If you need more visual debugging while tuning, connect the output to Show Image to inspect changes frame-by-frame.