Dehazing library is a standalone portable C++ dynamic library that provides instant enhancement of video streams or separated images that have been injured by atmospheric phenomena such as fog, mist, rain, dust, etc. It is fully automatic and extremely easy to use.
The dehazing method is based on an innovative and powerful combination of state of the art approaches to obtain haze-free images. In case of colour images, it offers near perfect colour quality in a wide range of conditions. Each pixel of an injured image is immediately corrected to acquire the perfect view.
The parallel data processing is highly optimized on all computation levels to maximally reduce the processing latency. Thanks to utilization of Open Computing Language framework (OpenCL), heterogenous parallel platforms are supported (CPU, GPU, DSP).
Easy to use
Due to its flexible software implementation and well documented API, the library can be easily integrated into other applications and video management systems. It is suitable for enhancement of both colour and grayscale images.
The dehazing library is especially unique for three reasons: the quality and versatility of its processing algorithms, its speed (real-time) and low memory requirements, and its utilization of common parallel proccessing platforms such as Nvidia or AMD graphics cards, and multicore CPUs.