CV2 Threshold Visualizer
An OpenCV desktop tool for comparing classic and adaptive thresholding on the same image.
- Project type
- Open Source
- Language
- Python
- Stars
- 1
- Forks
- 0
FIG. 01Original and thresholded images displayed side by side
Problem
Lighting and image detail affect which thresholding method works best. Seeing the effect of each parameter helps choose an appropriate preprocessing step.
Solution
I built a desktop tool for loading images and experimenting with classic and adaptive thresholding. The original and processed image appear side by side, with controls for threshold value, neighborhood size and the C constant.
Image processing features
- Classic thresholding: Binary, Binary Inverse, Trunc, To Zero and To Zero Inverse.
- Adaptive thresholding: Mean and Gaussian methods for examining local lighting differences.
- Comparison: Adjust parameters, apply a method and inspect the output beside the original.
- Export: Save the processed image.
Technical structure
OpenCV and NumPy handle image processing, Tkinter provides the desktop interface and Pillow displays the images. The project makes classical computer vision techniques accessible without using a trained model.
Source code
Setup instructions and usage examples are available in the GitHub repository.