IRJET- A Comparative Analysis of various Visibility Enhancement Techniques through Single Image Defogging

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 10 | Oct 2018 www.irjet.net p-ISSN: 2395-0072
© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 784
A Comparative Analysis of Various Visibility Enhancement Techniques
through Single Image Defogging
Suganiya.S1, Shantha Preetha.S2, Swathika.M3
1,2,3Department of Electronics and Communication Engineering, Bharathiyar College of Engineering and
Technology, Karaikal, Tamil Nadu, India
----------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - This paper reports various fog removal
algorithms how fog is removed and improved in visibility are
reviewed. While capturing an image in foggy weather
condition degrades the image due to the presence of airlight.
The airlight in each region of the degraded image is detected
and restored. Which states that the fog formation is the
function of the depth and the depth information is estimated
with various assumptions or prior information. The defogging
technique entangled problems related to visual surveillance,
intelligent vehicle, tracking and remote sensing. The overall
objective of this paper is to analyse the various techniques for
efficiently eliminating the fog from the digital image.
Key Words: Comparative study,airlight,Fogremoval,Image
enhancement, Performance evaluation.
1. INTRODUCTION
The main objective of image processing is to understand,
recognize and interpret thedata fromtheimagepattern.And
in some cases, due to the presence of moistureless particles
such as dust, smoke, snow, haze and fog corrupt the image.
The tiny water droplets suspended in the air cause fog andit
is generally classified according to the physical process
producing saturation or near-saturationoftheair.The water
droplet causes absorption and scattering leadtoattenuation
(reduces the contrast) and airlight (whiteness effect). When
a light from the scene comes towards the camera or the
observer gets attenuated due to scattering through water
droplets and degrades the image quality.Theseaerosolslack
the sensitivity of cameras and the human eye to resolve so
an effective fog removal techniqueisenhancedtorestore the
degraded image.
Looking from the atmospheric point of view, weather
conditions get varied regarding size and type of the particle
present in the air. Bad weather condition is broadly
categorised as steady and dynamic by the type of the visual
effect. In steady lousy weather, the water droplet is tiny and
it is steady floating in the air which includes fog, mist, and
haze. Where in the case of dynamic weather, the water
droplet is found to have 1000 times larger volume than
steady weather which includes snow andrain.Radiationfog,
Advection fog Upslope fog, Ice fog, Freezing fog, Fog mixing
fog, Frontal fog are main types of fog.
The fog removal technique is broadly classified into image
enhancement and image restoration. The primary objective
of image enhancement is processing the given image and
producing a more suitable image whereas in case of image
restoration the corrupted or defogged image is processed to
evaluate the original image.
The contrast and colour characters of the image degrade
drastically under foggy weather condition. Generally, clear
day images have more contrast than the foggy image so fog
removal algorithm isdesignedtoenhancethescenecontrast.
Which tends to be challenging as the recovery of luminance
and chrominance and maintaining the colour fidelity is
complex. Pixel value should be taken into consideration asit
gets saturated in case of over enhancement. So there are
some constraints taken into consideration for image
enhancement which are the preservation of appropriate
colour fidelity and to avoid saturation of an image.
Image restoration method is much better than the image
enhanced as it maintains the detailed information and also
produces a natural result. The restoration technique is
classified into multiple images and a single image defogging
method. In multiple imagesdefogging,twoormoreimagesof
the same scene are used which operated under weather
condition, polarization and depth map based techniques.
Whereas in the case of single image defogging, the single
input image is taken and this method depends upon
statistical assumption and essence of the scene. This paper
reports various fog removal algorithms how fog is removed
and improved in visibility are reviewed. The overall
objective of this paper is to analyse the various techniques
for efficiently eliminating the fog from a digital image.
2. LITERATURE REVIEW
[1] OPTIMAL TRANSMISSION ESTIMATION VIA FOG
DENSITY PERCEPTION FOR EFFICIENT SINGLE IMAGE
DEFOGGING
Here in this paper, a simple and practical prior constraint
single image defogging technique isadopted.Thefogdensity
of a recovered image is directly estimated rather than
approaching prior assumption. Primarily from the foggy
image, three fog relevant statistical features are derivedand
then this fog relevant features develop a simple fog density
evaluator (SFDE). This low computational evaluator could
precisely predict the fog density of a single image without a
reference of fog-free images. Secondarily a transmission and
fog density based formulation is developed via SFDE for the
given foggy image patch. Two optimal and effective
transmission map are derived for evaluating the
transmission value using the Optimal Transmission model
via SFDE (OTSFDE) and a Simpler Optimal Transmission
model via SFDE (SOTSFDE).

[2] CONTRAST IN HAZE REMOVAL: CONFIGURABLE
CONTRAST ENHANCEMENT MODEL BASED ON DARK
CHANNEL PRIOR
Conventionally, De-hazing is performed by adjusting the
contrast and saturation to improve the quality of the
reconstructed image. The difficulty in haze removal
algorithm is reformulated subjecting to luminance
reconstruction scheme based on statistical analysis of
luminance value. The augmentation of contrast is based on
the variance in the gradient space and the interpretation of
contrast shows that dark channel magnifies the diversity
details by maximizing the changesininputimagegradientor
the saturation of the scene radiance is enhanced by
minimizing the difference to the estimated initial dark
channel. The resultant contrast value supercilious for the
given brightness value.Here theatmospherelightestimation
module operates on colour constancy method which
outperforms even when noise is considered. And the
luminance-oriented optimized framework runs at a
processing time of 0.55 seconds for the 1-megapixel image.
[3] SINGLE IMAGE VISIBILITY RESTORATION USING DARK
CHANNEL PRIOR AND FUZZY LOGIC
In this paper, the fog removal algorithm is designedinsucha
way that it is time efficient and competent in both
homogeneous and heterogeneous. Here we have DCP (dark
channel prior) algorithm in which the prior assumes that a
fog-free, clear image has intensity value close to zero in any
one of the colour channel. The DCP algorithm is fused with
fuzzy contrast enhancement technique which converts the
image into the fuzzy domain and spatial operation for fast
and improved contrast while the Global Histogram
Equalization (GHE) or Adaptive Histogram Equalization
(AHE) over-saturate the patch. It also reinforces the minute
details of the degraded fog image.
[4] VISIBILITY ENHANCEMENT TECHNIQUE FOR HAZY
SCENES
Here in this paper, an effective visibility enhancement
technique for single image de-hazing is designed using Dark
Channel Prior technique.Estimatethedark pixelshavinglow
intensity at any one of the RGB channels and this dark
channel provides exact estimation for obtaining the
transmission map. Fortheedgepreservationoftransmission
map, the bilateral filter is used. To obtain the gamma
correction technique and for estimating the exact colour of
the hazy input image, a Laplacian distribution value is used.
And for evaluating the sufficient transmission map, gamma
correction technique is used. To evaluate the quality of the
enhanced image, performance metric such asPSNR,emetric
and σ metric are used for measurement.
[5] EFFECT OF VARIOUS MODEL PARAMETERS ON FOG
REMOVAL USING DARK CHANNEL PRIOR
Here in this paper, the dark channel prior technique is used
which evaluate the transmission map from the constant
transmission parameter, which ranges between 0.9 and 1.
The obtained output for different filter produces a varying
constant parameter and on changing image patch size the
diverse results are studied which shows that the increase in
constant transmission parameter leads to increase in the
contrast of the scene. It concludes when the patch size is
increased, the image quality increases and noise decreases.
[6] DEHAZING FOR IMAGE AND VIDEO USING GUIDED
FILTER
A high-performance visionalgorithmisrequiredforeffective
haze removing. So here in this paper, a fast real-time image
and video dehazing method are proposed. And also the
airlight and the down-sampled transmission estimationand
extraction are performed ease using this proposed
algorithm. The improved guided filter isusedtoestimate the
transmission map which can further refined and up-
sampled. The obtained results show the algorithm
outperforms regarding speed and ability to improve
visibility.
[7] FOG DETECTION FOR DE-FOGGING OF ROAD DRIVING
IMAGES
In this paper generally, the fog removal technique
deteriorate the visual due to excessive contrast
improvement. Here the fog detection algorithm is designed
such that it selectively apply de-fogging method only at a
foggy region. Besides, an excessive contrast enhancement
adjustment and luminance compensation are done to avoid
too dark output. This proposed algorithm produces 97% of
fog detection accuracy and the subjective image quality is
improved.
[8] IMAGE HAZE REMOVAL USING IMAGE VISIBILITY
RESTORATION (IVR) & EDGE PRESERVING
DECOMPOSITION (EPD)
In this paper, the removal of haze is done by using the dark
channel prior algorithm and the estimation of atmospheric
light technique. To obtainthetransientimage,thepixel value
in dark region and atmospheric variation is estimated. If the
size of the image increases, it estimates that PSNR quality is
low and the computational time is high. If the size of the
image decreases, it estimates that PSNR quality is high and
the computational time is low. As a result, the high-quality
haze-free image is obtained.
[9] ACCELERATED FOG REMOVALFROMREALIMAGESFOR
CAR DETECTION
In this paper, an accelerated image enhancement technique
is used to detect the number of cars for traffic management.
Fog removal is based on simplified dark channel priorwitha
combined filter. The combinedfilterconsistsoftheproposed
adaptive filter with edge preserving technique to modify
transmission map promptly. Finally, the car detection
algorithm is used to decide the presence or absence of a car
in each image. The computational time of the proposed
algorithm is low.

[10] VISIBILITY ENHANCEMENT WITH SINGLE IMAGE FOG
REMOVAL SCHEME USING A POST-PROCESSING
TECHNIQUE
In this paper, visibility enhancement is done by effective
post-processing technique using single grey or a colour
image through fog removal. The dark channel prior
algorithm is used for improving visibility. This algorithm
preserves sharp details of the defogged image and also
maintains the colour quality of the defogged image. WLS
filter is used in post-processing technique. Comparatively,
single image processing in the proposed system is better
than an existing system.
[11]VISIBILITY ENHANCEMENT OF REAL-TIME FOGGY
VIDEOS
In this paper, a new and simple visibility enhancement
method is proposed which is implemented in the Graphics
processing unit (GPU) in real time. Kalman filter is used to
reduce the processing time for video frames and it is also
observed that the average processing time for SD video
stream is 5ms in which it is superior to the other observed
implementations. They can extend our method for better
restoration using anisotropic diffusion implementedinGPU.
[12] IMAGE-BASED AUTOMATED HAZE REMOVAL USING
DARK CHANNEL PRIOR
Previously a dehazing mechanism was developed based on
dark channel prior which cannot automatically set thepatch
size and the sky regions transmission value. The current
paper tries to fill this gap to automate these values. In this
paper, they proposed a practical algorithm for haze removal
focusing on the removal of significant demerits remained in
previous works. It mainly uses the concept of the dark
channel prior and proposed some set of assumptionstogeta
better result. Also, adaptive result calculation is the central
theme of our work. Simulation is done by taking around 50
natural hazy images. From both subjective and objective
measures, our methodgives betterresultscomparedtosome
existing methods.
[13] INCREASE DEHAZING PROCESS USING FAST GUIDED
FILTER ON THE DARK CHANNEL PRIOR
In this paper, haze removal is done by using a guided filter
and fast guided filter on the dark channel prior. Execution of
fast guided filter in the dark channel prior is faster than
guided filter implementation in the dark channel prior. The
resultant image is separated by the effect of fog in a better
quantity.
[14] A NOVEL IMAGE DEFOGGING ALGORITHM BASED ON
MULTI-RESOLUTION FUSION TRANSFORM
In this paper, they propose a novel algorithm based on a
fusion model integrated with a multi-resolution
approximation technique. They present a multi-resolution
defogging algorithm for extracting foreground objects of
interest from weather degraded images and enhancing the
extracted regions visibility at the same time. This method
yields accurate results and faster than existing de-hazing
strategies.PSNRismaximizedandcomputational complexity
is reduced.
[15]SINGLE FOG IMAGE RESTORATION VIA MULTI-SCALE
IMAGE FUSION
In traditional prior methods, have an issue on halo artifacts
and brightness distortion so to overcome thistheyproposed
an algorithm based on the multi-scale fusion of single image
restoration. The entire regionisdividedintotworegions, the
global atmospheric light can be effectively obtained in the
sky regions. The new Kirsch operator with adaptive
boundary constraint designed to optimize the transmission.
From the experimental results, it is observed that the
method outperforms regarding both efficiency and the
dehazing visual effect.
[16]HAZE REMOVAL USING THE DIFFERENCE-STRUCTURE
PRESERVATION PRIOR
Here in this paper, the dehazing algorithm designed on the
basis of difference structure-preservationprior,whichcould
estimate the optimal transmission map and restores the
actual scene. In order toobtaina moreaccuratetransmission
map, an assumption is made that an image patch is
approximated by a spare linear combination of an element
from a neighbour basis set. Here thesimilarstructureisused
throughout as possible and the difference between similar
patches are maintained. So as the result the highest SSIMs
(structural similarity image) is achieved as the structural
consistency is retained throughout the dynamic difference-
structure-preservation process.
[17] IMAGE DEHAZING USING NON-SYMMETRY AND ANTI-
PACKING MODEL BASED ON DARK CHANNEL PRIOR
Here in this paper, a novel method is added along with the
dark channel prior algorithm based on the Non-symmetry
and Anti-packing Model (NAM). Also, an auto level is used to
enhance the haze-free image's visual effect. TheNAMisused
to calculate the atmospheric light and the guided filter is
used to estimate the accurate transmission. Also, this
method shows several advantages of the NAM when
compared to a quadtree. Primarily, the blocks of NAM are
rectangular in structure and the size gets varied to avoid
segmenting the image into smaller blocks. So as the result
the NAM operates faster than the quadtree. Secondarily, the
each NAM block is standardized providing accurate
atmospheric light.
[18] A SYSTEM ARCHITECTURE FOR REAL TIME TRAFFIC
MONITORING IN FOGGY VIDEO
Here, this paper presents an architectureforreal-timetraffic
monitoring systems and it is required to satisfy two
significant constraints. Primarily,thedefoggedimageshould
be quality enough for further processing such as tracking
and object detection. Secondarily, the proposed algorithm
should be computationally cheap for real-time processing.
The proposed paper consist of an N thread for real-time
monitoring and the parallel architecture provides reduces
the processing time. The experimental result shows the
output obtained is suitable for live fog removal.
[19] IMAGE DEHAZING BASED ON REGION GROWING
The conventional single image haze removal algorithm has
an issue in error in atmospheric light value, the lower

transmission in the sky area and the problem of halo and
splashes artifact. Here an improved weight-based quad-tree
hierarchical search algorithm to better select atmospheric
light A and the seed point of region growing. There are three
conditions taken into consideration to calculate the
transmission, where the higher is contrast, the lower
information loss, the more balance histogram of the haze-
free image. For refine transmission, the edge preserving
filter is applied and down-sampling is done to achieve the
proper result.
[20] A FAST METHOD OF FOG AND HAZE REMOVAL
The primary objective of this paper is to enhance the
visibility, saturation, contrast and reduce noise in the foggy
image. Here they have introduced a method that uses the
single frame for enhancing foggy images using multilevel
transmission map. In comparison, this technique is fast and
free from noise or artifacts generated while processing
enhancement techniques. It is observedthatthetechniqueis
suitable for VGA resolution and it shows betterperformance
regarding both processing time and quality.
[21] SINGLE IMAGE FOG REMOVAL ALGORITHM BASEDON
AN IMPROVED DARK CHANNEL PRIOR METHOD
Here they have proposed a fast single image fog removal
algorithm based on an improved dark channel prior.Anditis
observed that the proposed algorithm can increase28.5% of
computing speed and 41.8% of image contrast ratio to the
conventional one. This algorithm can even remove fog
efficiently without the influence in the night too. And this
algorithm is suitable for the surveillance system and real-
time computing in an embedded system.
[22] MODIFIED DARK CHANNEL PRIOR MODEL AND
GAUSSIAN LAPLACIAN FILTERING WITH TRANSMISSION
MAP FOR FOG REMOVAL
The proposed system is a modified dark channel prior and
Gaussian Laplacian filtering (GLP) with transmissionmap in
which the GLP is used to remove the noise from the fog
image after that matte for recovering the fog-free image.
From the results, it is observed that this algorithm can
outperform as edge preservation smoothing approach has
provided quite promising results regarding peak signal to
noise ratio (PSNR) and entropy and execution time.
[23] A NEW FAST METHOD FOR FOGGY IMAGE
ENHANCEMENT
Here a novel method is proposed to enhance the contrast in
foggy images and this develops an imageatmosphericmodel
which is based on the Koschmieder's theory of atmospheric
vision. To achieve an outline of a strength of the fog in
different areas morphological operators operation is
performed. This proposed algorithm outperformsregarding
quantitative and qualitative analysis and also the
computation time is low.
[24] VECTORIZATION AND OPTIMIZATION OF FOG
REMOVAL ALGORITHM
This paper proposedtoapproachvectorization,optimization
and low memory capacity. An optimized anisotropic
diffusion, histogram stretching and smoothing based fog
removal algorithm is proposed. 70% of the time complexity
is eliminated using anisotropic diffusion and the accuracy is
achieved using optimization technique but it is neglectedfor
significant improvement. Heretheperformancedefoggingof
the algorithm is increased up to 90 fps (approx.) for VGA
image on DSP platform.
[25] VISIBILITY ENHANCEMENT THROUGH SINGLE IMAGE
FOG REMOVAL
Here a novel and effective algorithm are proposed for single
image fog removal that is capable of handling images of gray
and colour channel. Weighted Least Square (WLS) and High
Dynamic Range (HDR) algorithm is fused with dark channel
prior. From the simulation results, it is observed that the
output frog-free image contains more clear edges with
details and better contrast. The primitive advantage of this
algorithm produces a high-quality image and also maintains
the colour quality.
[26] A HAZE DENSITY AWARE ADAPTIVE PERCEPTUAL
SINGLE IMAGE HAZE REMOVAL ALGORITHM
In conventional methods usually, require complicated
manual parameters setting to the variance of input. Dark
channel prior is considered to be the most efficient de-haze
method but there is some limitation including low
luminance, sky region distortion and low saturation are
inevitable. Here, this paper introduces a haze density
detector which adaptively adjusts the parameter settings
and besides, it improves the original dimrecoveredimageby
adaptively adjusting exposure and colour saturation YCbCr
colour space. Furthermore, a fastguidedfilterisemployed to
refine the transmission map and the experimental results
show that the proposed method outperforms both
objectively and subjectively.
[27]DEVELOPMENT OF IMAGE DEHAZING SYSTEM
Here in this paper, a mean channel guided algorithm for
defogging is presented whose function is more accurate and
robust as compared withthe conventionmethodologies.And
it is hardware-implemented version will work on low cost,
low power and a portable processingcoreraspberrypialong
with a display screen. The obtained results of mean channel
guided prior are compared with DCP, MCP, SIFRGMF. The
obtained visibility enhancement algorithm performs
qualitatively and quantitatively better and efficient in
removing haze from synthetic as well as real-life images.
[28] IMAGE DEHAZING USING DARK CHANNEL PRIOR AND
THE CORRECTED TRANSMISSION MAP
Single image de-hazing based on dark channel prior may
encounter colour distortion in a bright region so to
overcome this situation 3 methods are proposed in this
paper. Primarilythetransmissionthresholdwasdetermined,
then by using the threshold to correct the transmission map
in different ways and make adaptive to fog, three algorithms

can effectively deal with the sky, white object and so on. The
complexity of calculation to refine the transmission map is
minimized using the fast guided filter. The experimental
result shows these methods are feasible to eliminate colour
distortion of out-door image and visibility is also enhanced.
[29] SINGLE IMAGE DEHAZING BASED ON ONE
DIMENSIONAL LINEAR FILTERING AND ADAPTIVE
HISTOGRAM EQUALIZATION METHOD
This paper presents a single image de-hazing method which
is based on a one-dimensional linear filter. The primary
objective of this paper is to resolve any type of foggy
problem by using this algorithm, based on mean
enhancement methodology and adaptive histogram
equalization method. YCbCr model excels in colour
compression in which Y luminance can be used separately
for storage in high resolution and the chromaticity
components treated separately to enhance the results.
Eventually, it achieves the linear complexity and results
demonstrates the effectiveness of the proposed algorithm.
[30] PARALLEL IMAGE DEHAZING ALGORITHM BASED ON
GPU USING FUZZY SYSTEM AND HYBRID EVOLUTION
ALGORITHM
Here in this paper, a parallel hybrid evolution algorithm
based on GPU is proposed to enhance the computational
performance. In conventional evolution algorithm, the
calculation of fitness function occupies most of the
computation time. So to overcome these circumstances we
implement this part on GPU by using CUDA framework to
reduce the computational load.Theexperimentresultsshow
that the algorithm proposed can remove the haze efficiently
and successfully.
3. GAPS IN LITERATURE SURVEY
Digital defogging algorithm plays an essential role in
numerous vision applications and it is found that the current
analysis mistreated numerous subjects. Limitations in the
literature survey are list below,
1. It is found that most of the discussedalgorithmhave
ignored the actual use of soft computing techniques
to improve the adaptively of the digital defogging
removal algorithm.
2. Majority of the paper has ignored the issue of
irregular light.
3. 85% of the existing methods have taken static
restoration value.
4. FUTURE WORK
So in near future, the problem of uneven illumination of the
digital fog removal has to be sorted out. To enhance the
visibility of image causedbyatmospheresuspendedparticles
like dust, haze and fog which causes failure in image
processing such as video surveillance systems, obstacle
detection systems, outdoor object recognition systems and
intelligent transportation systems. And visibility restoration
techniques should be developed to run under various
weather conditions.
5. CONCLUSION
This paper investigates various fog removal techniques
described here, the majority of the scientific study has
ignored several issues i.e., no techniqueisbetterfordifferent
kind of circumstances. The effectiveness of the methods,
different qualitative assessment are evaluated and the
experimental results demonstrate the used methods show
good results for fog degraded visuals. This analysis
contributes to developing a new and better fog removal
algorithm.
ACKNOWLEDGEMENT
We would like to thank our HOD, Dr. J. Samuel Manoharan
for giving us too much good knowledge, experience and
support.
REFERENCES
[1] Zhigang Ling, Jianwei Gong, Guoliang Fan, “Optimal
transmission estimation via fog density perception for
efficient single image defogging” in IEEE Transactions
on Multimedia, Aug 2015, pp. 1-13.
[2] Ping-Juei Liu, Shi-Jinn Horng*, Jzau-Sheng Lin, and
Tianrui Li, “Contrast in haze removal: configurable
contrast enhancement model based on dark channel
prior” in IEEE Transactions on Image Processing, 2018,
pp. 1-16.
[3] Subhadeep Koley, Ahana Sadhu, Hiranmoy Roy,
soumyadip Dhar, “Single image visibility restoration
using dark channel prior and fuzzy logic” in IEEE
International Conference on Electronics, Materials
Engineering & Nano-Technology (IEMEN Tech), 2018,
pp. 1-7.
[4] R.Ahila Priyadharshni, S.Aruna, “Visibility enhancement
technique for hazy scenes” in IEEE International
Conference on Electrical Energy Systems (ICEES),2018,
pp. 1-6.
[5] Neha, Rajesh Kumar Aggarwal, “Effect of various model
parameters on fog removal using dark channel prior” in
IEEE International Conference On Recent Trends in
Electronics Information & Communication Technology
(RTEICT), May 2017, pp. 1-4.
[6] Zheqi Lin, Xuansheng Wang, “Dehazing for image and
video using guided filter” in Open Journal of Applied
Sciences, World Congress on Engineering and
Technology, 2012, pp. 1-5.

[7] Kwang Yeon Choi, Kyeong Min Jeong, ByungCheol Song,
“Fog detection for de-fogging of road driving images” in
IEEE International Conference on Intelligent
Transportation Systems (ITSC), 2017, pp. 1-6.
[8] Zinab Khan, Dr.Anjali Asish Potnis, “Image haze removal
using image visibility restoration (IVR) & edge
preserving decomposition (EPD)” in IEEE International
Conference on Computer Communication and
Informatics (ICCCI -2017), Jan 2017, pp. 1-5.
[9] Rawan Younis, Nabil Bastaki, “Accelerated fog removal
from real images for car detection” in IEEE-GCC
Conference and Exhibition (GCCCE), 2017, pp. 1-6.
[10] Md. Imtiyaz Anwar, Arun Khosla, and Gajendra Singh
“Visibility enhancement with single image fog removal
scheme using a post-processing technique” in IEEE
International Conference on Signal Processing and
Integrated Networks (SPIN), 2017, pp. 1-6.
[11] Manvendra Singh Chauhan, Jayashree Pradhan,
Pradipta Roy, Dipak Das,“Visibilityenhancementof real-
time foggy videos” in IEEE International Conference on
Image Information Processing (ICIIP), 2017, pp. 1-5.
[12] Mohammad Shorif Uddin∗, Bishal Gautam∗, Aditi
Sarker∗, Morium Akter∗ and Mohammad Reduanul
Haque†, “Image-based automated haze removal using
dark channel prior” in IEEE Region 10 Humanitarian
Technology Conference (R10-HTC), Dec 2017, pp. 1-4.
[13] Rizal Mutaqin, Fresy Nugroho, Nugraha Gumilar,
“Increase dehazing process using fast guided filter on
the dark channel prior”inIEEEInternational Conference
on Electrical, Electronics and Information Engineering
(ICEEIE), 2017, pp. 1-6.
[14] Zhuohan Cheng, Xin Xiang, Yaochen Shen, “A novel
image defogging algorithm based on multi-resolution
fusion transform” in IEEE International Conference on
Opto-Electronic Information Processing, 2017, pp. 1-4.
[15] Yin Gao, Yijing Su, Qiming Li, Jun Li*, “Single fog image
restoration via multi-scale image fusion” in IEEE
International Conference on Computer and
Communications, 2017, pp. 1-6.
[16] Linyuan He, Jizhong Zhao, “Haze removal using the
difference-structure preservation prior” in IEEE
Transactions on Image Processing, 2016, pp. 1-13.
[17] Yunping Zheng, Yukang shu, changting cai, “Image
dehazing using non-symmetry and anti-packing model
based on dark channel prior” in International
Conference on Natural Computation,FuzzySystemsand
Knowledge Discovery (ICNC-FSKD 2017), 2017, pp.1-6.
[18] Sangkyoon Kim, Soonyoung Park, Kyoungho Choi, “A
system architecture for real time traffic monitoring in
foggy video” in IEEE Korea-Japan Joint Workshop on
Frontiers of Computer Vision (FCV), 2015, pp. 1-4.
[19] Wei Liu, Ping Ye, Hanxu Sun, “Image dehazing based on
region growing” in IEEE International Conference on
Systems and Informatics (ICSAI), 2017, pp. 1-6.
[20] Veeranjaneyulu Toka, Nandan Hosagrahara
Sankaramurthy, Ravi Prasad Mohan Kini, Prasanna
Kumar Avanigadda, sibsambhu kar, “A fast method of
fog and haze removal” in IEEE International Conference
on Acoustics, Speech and Signal Processing (ICASSP),
2016, pp. 1-5.
[21] Chi-Chia Sun, Heng-Chi Lai, Ming-Hwa Sheu, Yi-Hsing
Huang, “Single image fog removal algorithmbasedonan
improved dark channel prior method” in IEEE
International Symposium on Intelligent Signal
Processing and Communication Systems (ISPACS), Oct
2016, pp. 1-4.
[22] Jaiveer Singh Sikarwar, Abhinav Vidwans, “Modified
dark channel prior model and gaussian laplacian
filtering with transmission map for fog removal”inIEEE
International Conference on Electrical, Electronics, and
Optimization Techniques (ICEEOT), 2016, pp. 1-6.
[23] Mohammad Javad Abbaspour, Mehran Yazdi,
Mohammadali Masnadi-shirazi, “A new fast method for
foggy image enhancement” in Iranian Conference on
Electrical Engineering (ICEE), 2016, pp. 1-5.
[24] Krishna Swaroop Gautam, Abhishek Kumar Tripathi,
M.V. Srinivasa Rao, “vectorization and optimization of
fog removal algorithm”inIEEEInternational Conference
on Advanced Computing, 2016, pp. 1-6.
[25] Md. Imtiyaz Anwar*, Arun Khosla, “Visibility
enhancement through single image fog removal” in
International Journal, Engineering Science and
Technology, 2017, pp. 1-9.
[26] Chuanzi He, Chendi Zhang, QingrongCheng, Xixiaoyi Jin,
Jianjun Yin “A haze density aware adaptive perceptual
single image haze removal algorithm” in Proceedings of
the IEEE International Conference on Information and
Automation Ningbo, Aug 2016, pp. 1-6.

[27] Amruta Deshmukh, Satbir Singh, “Development of
image dehazing system” in IEEE International
Conference on Wireless Networks and Embedded
Systems(WECON), 2016, pp. 1-5.
[28] Lei Shi, Li Yang, Xiao Cui, Zhigang Gai, Shibo Chu, jing
Shi, “Image dehazing using dark channel prior and the
corrected transmission map” in IEEE International
Conference on Control, Automation and Robotics, 2016,
pp. 1-4.
[29] Ashok Shrivastava, Dr Sanjay Jain, “Single image
dehazing based on one dimensional LinearFilteringand
Adoptive Histogram Equalization method” in IEEE
International Conference on Electrical, Electronics, and
Optimization Techniques (ICEEOT), 2016, pp. 1-5.
[30] Che-Lun Hung, Hsiao-Hsi Wang, Ren-You Yan, “Parallel
image dehazing algorithm based on GPU using fuzzy
system and hybrid evolution algorithm” in IEEE/ACIS
International Conference on Software Engineering,
Artificial Intelligence, Networking and
Parallel/Distributed Computing (SNPD), 2016, pp. 1-3.
SUGANIYA. S doing final year in
B.Tech from BHARATHIAYR
COLLEGE OF ENGINEERING AND
TECHNOLOGY, India.
SHANTHA PREETHA.Sdoingfinal
year in B.Tech from
BHARATHIAYR COLLEGE OF
ENGINEERINGANDTECHNOLOGY.
SWATHIKA. S doing final year in
B.Tech from BHARATHIAYR
COLLEGE OF ENGINEERING AND
TECHNOLOGY, India.
AUTHORS

IRJET- A Comparative Analysis of various Visibility Enhancement Techniques through Single Image Defogging

More Related Content

What's hot (20)

Similar to IRJET- A Comparative Analysis of various Visibility Enhancement Techniques through Single Image Defogging (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET- A Comparative Analysis of various Visibility Enhancement Techniques through Single Image Defogging