Practical Computation Acceleration for Large Size Single Image Filtering Based Haze Removal

The main defect of traditional filtering based image dehazing approaches is huge time consuming, especially for large size image haze removal. In the work, a serial of practical computation acceleration strategies for bilateral filtering based haze removal are proposed. Our approach strength lies in its wide practicability: better dehazing with very fast computation speed even for large size images. Since our approach requires less time consuming corresponding to less CPU load, it will be promising for some real time application, such as smart TV image enhancement.

[1]  Yoav Y. Schechner,et al.  Advanced visibility improvement based on polarization filtered images , 2005, SPIE Optics + Photonics.

[2]  Chunxia Xiao,et al.  Fast image dehazing using guided joint bilateral filter , 2012, The Visual Computer.

[3]  Jean-Philippe Tarel,et al.  Fast visibility restoration from a single color or gray level image , 2009, 2009 IEEE 12th International Conference on Computer Vision.

[4]  Jian Sun,et al.  Single image haze removal using dark channel prior , 2009, 2009 IEEE Conference on Computer Vision and Pattern Recognition.

[5]  Narendra Ahuja,et al.  Real-time O(1) bilateral filtering , 2009, CVPR.

[6]  Pierre Kornprobst,et al.  Bilateral Filtering , 2009 .

[7]  Roberto Manduchi,et al.  Bilateral filtering for gray and color images , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).