Singh and Kumar [1] have proved that the dark channel prior (DCP) is an adequate haze removal model. However, its procedure causes annoying halo and gradient reversal artifacts. Also, the existing approaches still suffer from color distortion, edge degradation, and halo artifacts issues. To overcome these problems, a fog removal technique based on modified gain coefficient filter (GCF) is implemented [2]. Also, modified restoration model (MRM) based DCP is proposed in [3] to minimize the color distortion rate. The partial differential equation-based filter has been designed to remove the fog from images in an efficient manner [4]. In [5], the transmission map obtained from DCP has been refined by designing an efficient gain coefficient-based trilateral filter (GGF). A novel integrated channel prior (ICP) is implemented in [1] to solve the sky-region problem associated with DCP. The gain intervention filter is also utilized to improve computational speed and edge preservation. The notch filter (NF) is proposed to improve the texture information of restored images [6]. However, these techniques [1–6] can be further improved by designing an efficient channel prior to deal with dense fog image. We attempt to restore the visibility of foggy outdoor images and concurrently suppress artifacts for restoration of radiometric detail in inclement weather circumstances. The major contributions of this study are: (1) A modified restoration method with dynamic threshold has been developed to resolve the sky region problem associated with DCP. The use of dynamic threshold value also reduces the color distortion rate. (2) The gradient guided image filter has been modified by considering the improved guide image (Gd). It has an ability to overcome the issue of halo and gradient reversal artifacts problems. The gain coefficientbased gradient guided image filter is then utilized to improve the coarse estimated atmospheric veil. Fog formation model. The fog formation model is mathematically evaluated as
[1]
Vijay Kumar,et al.
Dehazing of outdoor images using notch based integral guided filter
,
2018,
Multimedia Tools and Applications.
[2]
Dilbag Singh,et al.
Single image haze removal using integrated dark and bright channel prior
,
2018
.
[3]
Vijay Kumar,et al.
Modified gain intervention filter based dehazing technique
,
2017
.
[4]
Vijay Kumar,et al.
Dehazing of remote sensing images using fourth-order partial differential equations based trilateral filter
,
2017,
IET Comput. Vis..
[5]
Vijay Kumar,et al.
Dehazing of remote sensing images using improved restoration model based dark channel prior
,
2017
.
[6]
Vijay Kumar,et al.
Defogging of road images using gain coefficient-based trilateral filter
,
2018,
J. Electronic Imaging.