# discocal

**Repository Path**: xiaolongshiyl/discocal

## Basic Information

- **Project Name**: discocal
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: detector
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2024-07-30
- **Last Updated**: 2024-07-30

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Unbiased Estimator for Distorted Conics in Camera Calibration (CVPR24)
## Description
<img src="./Figs/overview.png" width="600" height="300">


As shown above, the circle center is not projected to the centroid of the distorted ellipse under perspective transformation and distortion. Without considering geometery of the distorted ellipse, existing circular pattern-based calibration methods are biased, which leads low calibration accuracy than a checkerboard pattern. 

**Our unbiased estimator completes the missing piece in the conic-based calibration pipeline and outperforms the checkerboard pattern-based calibration.**

## Camera model
We assume pin-hole camera model with radial distortion.
<img src="./Figs/camera_model.png" width="600" height="200">

Calibration results:	
	fx, fy, cx, cy, skew, d1, d2, ... dn



# How to use
## Dependency
- [Ceres-Solver](http://ceres-solver.org/index.html)
- [Eigen3](https://eigen.tuxfamily.org/dox/index.html)
- opencv4

We also furnish official docker image.

	docker pull chaehyeonsong/discocal:latest


## How to run
	## Build
	cd [your path]/discocal
	mkdir build
	cd build
	cmake ..
	make

	## Run
	./main.out [n_x] [n_y] [n_d] [img_dir_path] [r(m)] [distance(m)] [0:rgb or 1:thermal]
	(ex) ./main.out 4 3 3 ../imgs/ 0.035 0.05 0

## Important: Circle detection
**To get high-quality results, plz check all pixels in the circle are correctly detected like this.**
<!-- ![sample](./Figs/detection_sample.png){: width="100" height="100"} -->
<img src="./Figs/detection_sample.png" width="400" height="300">

If you don’t want to verify images, turn off the “check_detection_results” option in "main.cpp".


Our circle detector finds the elliptical blobs in the image with the color value and ellipse test.
#### Parameters:
- **color_threshold_max**: the intensity threshold between the blob and surrounding pixels (intensity of a blob pixel < color_threshold_max)
- **size_threshold**: the number of pixels composing a blob is bigger than this value
- **fullfill_threshold**: the difference between real area and estimated area resulting from ellipse fitting
- **eccentricity_threshold**: the length ratio between a blob's major and minor axis.
You can refine these parameters in the TargetDetector class.




# Bibtex
	@INPROCEEDINGS { chsong-2024-cvpr,
		AUTHOR = { Chaehyeon Song and Jaeho Shin and Myung-Hwan Jeon and Jongwoo Lim and Ayoung Kim },
		TITLE = { Unbiased Estimator for Distorted Conic in Camera Calibration },
		BOOKTITLE = { IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) },
		YEAR = { 2024 },
		MONTH = { June. },
		ADDRESS = { Seattle },
	}