🟧 Image Histogram Equalization

Function Description

The operator adjusts the grayscale histogram of the image to adjust the distribution of pixels with different brightness in the image, making originally concentrated pixels in certain brightness intervals more dispersed, thereby making the overall contrast of the image more obvious.

Usage Scenarios

Used as image preprocessing before feature detection, edge detection and other algorithms to enhance feature prominence and improve robustness of subsequent algorithms.

Input Output

Input

Image: Input images or image lists that need histogram equalization.

Output

Result Image: Images or image lists after equalization processing.

Parameter Description

Global Equalization

Parameter Description

Transform based on the grayscale histogram of the entire image, applying a unified contrast enhancement standard to the entire image.

Parameter Adjustment

This method is simple and fast, suitable for scenarios that need to improve overall contrast. However, if the image contains large areas of extremely bright or extremely dark regions, it may cause local detail loss or noise amplification.

Parameter Range

[0, 100], Default value: 2

Contrast Adaptive Equalization

Parameter Description

Divide the image into several small rectangular regions and perform histogram equalization independently for each small region. This method can better enhance local details while suppressing excessive noise amplification through "contrast limiting".

Parameter Adjustment

This method performs better in enhancing local details and can effectively suppress noise. When the image has uneven lighting or needs to preserve fine textures, this method is recommended.

Contrast Limiting Threshold

  • Increasing this value: Allows higher contrast enhancement, image details will be more prominent, but may amplify noise.

  • Decreasing this value: Stronger noise suppression, image looks more natural, but contrast enhancement effect is relatively weaker.

Grid Size: Divide the image into multiple grids for local processing, for example, 8 means dividing the image into an 8x8 grid.

  • Smaller grid size: Image enhancement effect is closer to global histogram equalization, local contrast improvement is not obvious.

  • Larger grid size: Local details of the image will be better enhanced, but computational load will also increase accordingly, and may introduce some unnatural local effects.

Parameter Range

Contrast limiting threshold: [0,100], Default value 2

Grid size: [1,30], Default value 8