Researching

Journals
  • Advanced Photonics
  • Photonics Insights
  • Advanced Photonics Nexus
  • Photonics Research
  • Advanced Imaging
  • View All Journals
  • Chinese Optics Letters
  • High Power Laser Science and Engineering
    • Articles
  • Optics
  • Physics
  • Geography
  • View All Subjects
    • Conferences
  • CIOP
  • HPLSE
  • AP
  • View All Events
    • News
    About CLP
    Search by keywords or author
    Journals >Laser & Optoelectronics Progress >Volume 57 >Issue 6 >Page 060002 > Article
    • Laser & Optoelectronics Progress
    • Vol. 57, Issue 6, 060002 (2020)
    Review on Key Technologies of Target Exploration in Underwater Optical Images
    Sen Lin1、3、4 and Ying Zhao1、2、*
    Author Affiliations
  • 1School of Electronic and Information Engineering, Liaoning Technical University, Huludao, Liaoning 125105, China
  • 2Institute of Graduate, Liaoning Technical University, Huludao, Liaoning 125105, China
  • 3State Key Laboratory of Robotics, Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, Liaoning 110016, China
  • 4Institutes for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences, Shenyang, Liaoning 110016, China
    • show less
    DOI: 10.3788/LOP57.060002 Cite this Article Set citation alerts
    Sen Lin, Ying Zhao. Review on Key Technologies of Target Exploration in Underwater Optical Images[J]. Laser & Optoelectronics Progress, 2020, 57(6): 060002 Copy Citation Text show less
    Traditional methods of underwater image feature extraction
    Fig. 1. Traditional methods of underwater image feature extraction
    Download full size
    MethodPrincipleCharacteristicLimitation
    Histogram-basedImproving image contrast throughnonlinear stretching of pixel valuesBright color,real-timeLack of texture details,increasing noise
    Retinex-basedRemoving the influence of illuminatinglight from the image to obtain thereflective properties of the objectBright color,high contrast,more detail informationOver-enhancement for thebrighter patch, unsaturationfor darker region
    Fusion-basedCombining relevant informationfrom two or more images intoa single image, which is moreinformative than any of the inputsEffective colorcorrection,high contrastLess robust, affected byartificial illumination
    Deep learning-basedImitating the working of thehuman brain in processing data,and improving the quality of underwaterimage through network trainingEffective colorcorrection, robustcontrast stretchingHard to get ground-truthimages, time-consumingfor training
    Table 1. Comparison of underwater image enhancement methods
    MethodImprovementCharacteristic
    DCP-basedTang et al.[25]:the depth-of-field map is obtained bydisparity between bright and dark channels, so as toestimate the background color of water body andtransmittance map more accuratelyHigh real-time performance,more authentic colors,influence of artificial lightsources, and inadequaterobustness and adaptability
    Xu et al.[26]:estimation oftransmittance map of images byred channel algorithms
    Xie et al.[27]:using the relationship betweenthe wavelength of visible light and the scatteringcoefficient to obtain transmittance maps of each color channel
    Deeplearning-based[23]URCNN: a convolutional layer and ReLU are used to generatefeature maps, then the batch normalization is addedbetween convolutional layer and ReLU to speed up training process.This pattern is repeated until the transmission map is outputMore authentic colors,robustness, and beingtime-consuming for training
    UIRNet:the transmission map and the backgroundlight are estimated and computedby BL-Net and TM-Net, respectively
    WaterGAN[24]:taking in-air images anddepth maps as input and generatingcorresponding synthetic underwater images as output
    Table 2. Comparison of underwater scene coefficient estimation methods
    AuthorMethodAdvantage and limitation
    Salmanet al.[43]A hybrid approach involving GMM,optical flow, and deep R-CNN tofine-tune the categorization of fishHigher classification accuracy;requirement on relatively morecomputational resources
    Siddiquiet al.[44]A cross-layer pooling algorithm usinga pre-trained convolutional neuralnetwork as a generalized feature detectorNo need for a large amount of training data;high classification accuracy;requirement of extensive computations
    Sunet al.[45]A CNN knowledge transfer framework forunderwater object recognition andextracting discriminative featuresfrom relatively low contrast imagesHigh real-time performance;no need for a large amount of training data;lower robustness
    Chuanget al.[46]An underwater fish recognition frameworkthat consists of a fully unsupervisedfeature learning technique and anerror-resilient classifierSuccessfully handled data uncertaintyand class imbalance in practicalclassification applications;lower robustness
    Caoet al.[42]A method to combine CNN andhand-designed features toimprove classification performanceBeing robust, reducing featuredimensionality without decreasingclassification performance;contour extraction needing original video frames
    Table 3. Comparison of depth neural network for underwater target detection and recognition
    MethodImprovementPrincipleCharacteristic
    Optical flow-basedModified Lucas-Kanade opticalflow method based on pyramidhierarchy and affine transformationEstimating the positionby calculating thevelocity in twoconsecutive framesNeed of a largenumber ofprecise imagefeature points
    Mean shift-based1) Adaptive mean shift algorithmthrough color histogram based onbackground and target region;2) adaptive mean shift algorithmcombined with edge information;3) combined with color, texture,HOG features, and deformablemulti-core algorithmUsing the colorhistogram of the targetas the searchfeature, and iteratingthe mean shiftvector continuouslyBeing robust torange variations;being influencedby multi-objects
    CNT-based1) Fast-CNT: improving thecomputing performance byselecting adaptive K value andomitting background filter;2) improved Fast-CNT:extracting each region containinga moving target through Gaussianmixture modelTracking target throughtwo-layer forwardconvolution networkslearning image featuresNo need for off-linetraining which istime-consumingand requires lotsof training data
    Table 4. Comparison of underwater target tracking methods
    Abstract
    Get PDF(in Chinese)
    Figures&Tables (5)
    Equations (0)
    References (62)
    Cited By (4)
    Get Citation
    Copy Citation Text
    Sen Lin, Ying Zhao. Review on Key Technologies of Target Exploration in Underwater Optical Images[J]. Laser & Optoelectronics Progress, 2020, 57(6): 060002
    Download Citation
    Set citation alerts for the article
    Tools
    Share
    Set citation alerts for the article
    Save the article for my favorites
    Paper Information
    Recommended Topics
    laser devices and laser physics
    Lasers and Laser Optics
    Laser physics
    laser manufacturing
    Instrumentation, Measurement and Metrology