Visibility Restoration: A Systematic Review and Meta-Analysis

Ngo, Dat; Lee, Seungmin; Ngo, Tri Minh; Lee, Gi‐Dong; Kang, Bongsoon

doi:10.3390/s21082625

Cited by 14 publications

(13 citation statements)

References 126 publications

(351 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For this comparison, we consider the first three columns of the figure that depict real scenes degraded by mild, moderate, and dense haze. It should be noted that results of FFD, DCP, CAP, DehazeNet, and MSCNN in the second column are adopted from Ngo et al [ 13 ].…”

Section: Experimental Resultsmentioning

confidence: 99%

“…Hence, this subsection systematically reviews pertinent research and summarizes the major turning points. A systematic review and meta-analysis of this field is available in [ 13 ].…”

Section: Literature Reviewmentioning

confidence: 99%

“… Qualitative comparison of the proposed AUDS with state-of-the-art benchmark methods on different images. Results of FFD, DCP, CAP, DehazeNet, and MSCNN in the ( b ) column are adopted from Ngo et al [ 13 ]. …”

Section: Figurementioning

confidence: 99%

See 2 more Smart Citations

Automating a Dehazing System by Self-Calibrating on Haze Conditions

Ngo

Lee

et al. 2021

Sensors

Self Cite

View full text Add to dashboard Cite

Existing image dehazing algorithms typically rely on a two-stage procedure. The medium transmittance and lightness are estimated in the first stage, and the scene radiance is recovered in the second by applying the simplified Koschmieder model. However, this type of unconstrained dehazing is only applicable to hazy images, and leads to untoward artifacts in haze-free images. Moreover, no algorithm that can automatically detect the haze density and perform dehazing on an arbitrary image has been reported in the literature to date. Therefore, this paper presents an automated dehazing system capable of producing satisfactory results regardless of the presence of haze. In the proposed system, the input image simultaneously undergoes multiscale fusion-based dehazing and haze-density-estimating processes. A subsequent image blending step then judiciously combines the dehazed result with the original input based on the estimated haze density. Finally, tone remapping post-processes the blended result to satisfactorily restore the scene radiance quality. The self-calibration capability on haze conditions lies in using haze density estimate to jointly guide image blending and tone remapping processes. We performed extensive experiments to demonstrate the superiority of the proposed system over state-of-the-art benchmark methods.

show abstract

Section: Experimental Resultsmentioning

confidence: 99%

“…Hence, this subsection systematically reviews pertinent research and summarizes the major turning points. A systematic review and meta-analysis of this field is available in [ 13 ].…”

Section: Literature Reviewmentioning

confidence: 99%

See 1 more Smart Citation

Automating a Dehazing System by Self-Calibrating on Haze Conditions

Ngo

Lee

et al. 2021

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…On the other hand, edge devices can introduce computing delays because of hardware limitations compared with the cloud. Several approaches to accelerate DNN inference at the edge devices, including a hardware-based approach that uses hardware implementation to tackle this issue [16]. For example, Eyeriss v2 [17] designs an accelerator for DNNs on mobile devices, while Karras et al [18] employ an FPGA board to accelerate DNN inference.…”

Section: Related Workmentioning

confidence: 99%

Towards Edge Computing Using Early-Exit Convolutional Neural Networks

Pacheco,

Bochie,

Gilbert

et al. 2021

Information

View full text Add to dashboard Cite

In computer vision applications, mobile devices can transfer the inference of Convolutional Neural Networks (CNNs) to the cloud due to their computational restrictions. Nevertheless, besides introducing more network load concerning the cloud, this approach can make unfeasible applications that require low latency. A possible solution is to use CNNs with early exits at the network edge. These CNNs can pre-classify part of the samples in the intermediate layers based on a confidence criterion. Hence, the device sends to the cloud only samples that have not been satisfactorily classified. This work evaluates the performance of these CNNs at the computational edge, considering an object detection application. For this, we employ a MobiletNetV2 with early exits. The experiments show that the early classification can reduce the data load and the inference time without imposing losses to the application performance.

show abstract

“…Visibility is measured by trained professionals in the traditional way [12]. In recent years, various techniques for estimating visibility via images were developed [13,14]. The features, such as transmittance [15], entropy [16], contrast [17], and high-frequency information [18], are extracted from images and used to estimate visibility.…”

Section: -500mentioning

confidence: 99%

Using High-Frequency Information and RH to Estimate AQI Based on SVR

Liaw

Chen

2021

Sensors

View full text Add to dashboard Cite

The Environmental Protection Administration of Taiwan’s Executive Yuan has set up many air quality monitoring stations to monitor air pollution in the environment. The current weather forecast also includes information used to predict air pollution. Since air quality indicators have a considerable impact on people, the development of a simple, fast, and low-cost method to measure the AQI value is a worthy topic of research. In this study, a method was proposed to estimate AQI. Visibility had a clear positive relationship with AQI. When images and AQI were compared, it was easy to see that visibility decreased with the AQI value increase. Distance is the main factor affecting visibility, so measuring visibility with images has also become a research topic. Images with high and low PM2.5 concentrations were used to obtain regions of interest (RoI). The pixels in the RoI were calculated to obtain high-frequency information. The high-frequency information of RoI, RH, and true AQI was used for training via SVR, which was used to generate the model for AQI estimation. One year of experimental samples was collected for the experiment. Two indices were used to evaluate the performance of the proposed method. The results showed that the proposed method could be used to estimate AQI with acceptable performance in a simple, fast, and low-cost way.

show abstract

Visibility Restoration: A Systematic Review and Meta-Analysis

Cited by 14 publications

References 126 publications

Automating a Dehazing System by Self-Calibrating on Haze Conditions

Automating a Dehazing System by Self-Calibrating on Haze Conditions

Towards Edge Computing Using Early-Exit Convolutional Neural Networks

Using High-Frequency Information and RH to Estimate AQI Based on SVR

Contact Info

Product

Resources

About