Automatically enhance image brightness based on fast haze removal[1]. Support 24-bit RGB image and 8-bit gray image.

Development Environment

• Microsoft Visual Studio Community 2017
• Windows 10
• OpenCV 3.3.0
• x64

Getting Started

1. Open IBEABFHR.sln with Microsoft Visual Studio 2017;
2. Config OpenCV according this Install OpenCV with Visual Studio;
3. Press Ctrl+B to build program;
4. Press F5 to run program;

Run Time

resolution	type	time
1024x768	gray	8.77ms
1024x768	color	16.24ms
1920x1080	gray	22.61ms
1920x1080	color	40.60ms
4160x2340	gray	104.57ms
4160x2340	color	186.14ms

The times above are the average time for call brighten100 times. If you just call brighten one times, the run time will be larger, because the first time doing image invert operation consumes extra time for unknown reason.

Adjust Parameters

There has two variable parameters :

1. radius in step 3, as the box filter radius. This parameter should not too small, otherwise result image will has halo artifact.
2. p in step 5, which controlls the result image's brightness. Bigger is it, brighter is the result image.

Effect Display

Reference

[1] 刘倩, 陈茂银, 周东华. 基于单幅图像的快速去雾算法[C]//25th Chinese Control and Decision Conference, 2013: 3780-3785.