https://github.com/laulauthom/haar-cascade-nms

OpenCV Haar cascade with non-maxima suppression
https://github.com/laulauthom/haar-cascade-nms

haar-cascade object-detection object-localization opencv

Last synced: 28 days ago
JSON representation

OpenCV Haar cascade with non-maxima suppression

• Host: GitHub
• URL: https://github.com/laulauthom/haar-cascade-nms
• Owner: LauLauThom
• License: mit
• Created: 2020-09-08T10:04:56.000Z (about 4 years ago)
• Default Branch: master
• Last Pushed: 2021-12-21T11:42:46.000Z (almost 3 years ago)
• Last Synced: 2024-10-11T16:22:56.845Z (28 days ago)
• Topics: haar-cascade, object-detection, object-localization, opencv
• Language: Jupyter Notebook
• Homepage:
• Size: 368 KB
• Stars: 4
• Watchers: 2
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Changelog: CHANGELOG.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# OpenCV Haar cascade with Non-Maxima Suppression

This simple project provides a `Detector` class extending the opencv `CascadeDetector`class with a new function: `detectAndFilter`.  

This function performs the detection using the cascade and filter the detections based on the score and their overlap to yield the best detections that do not overlap above the selected `overlap-threshold`.

For more details about Non-Maxima Suppression, you can refer to my [article](https://doi.org/10.1186/s12859-020-3363-7) about Multi-Template Matching, which uses a similar strategy (or other online resources).  

I also described/experimented with cascade deectors as part of my [PhD thesis](https://doi.org/10.11588/heidok.00030804).  

Here the NMS shipped with OpenCV is used.

# Installation

Using pip:  

`pip install haar-cascade-nms`

Then to use in your scripts:  

```python

from haarcascade import Detector

myDetector = Detector(filepathToCascade)

myDetector.detectAndFilter(image, etc...)

```

See the tutorial notebook for a more detailed documentation.