Journals Articles Conferences News About CLP
Submit a paper Email Alert
Search
Search
Articles
Journals
News
Advanced Search
Top Searches
2D MaterialsTransformation opticsQuantum PhotonicsSmall lasersSemiconductor UV PhotonicsSupersymmetric microring laser arrays

Acta Optica Sinica, Volume. 39, Issue 12, 1210001(2019)

Semantic Segmentation of Remote Sensing Image Based on Neural Network

Ende Wang1,2,3, Kai Qi1,2,3,4、*, Xuepeng Li1,2,3, and Liangyu Peng1,2,3
Author Affiliations
  • 1Shenyang Institute of Automation, Chinese Academy of Sciences, Shenyang, Liaoning 110016, China
  • 2Institute for Robotics and Intelligent Manufacturing, Chinese Academy of Sciences, Shenyang, Liaoning 110169, China
  • 3Key Laboratory of Opto-Electronic Information Processing, Chinese Academy of Sciences, Shenyang, Liaoning 110016, China
  • 4College of Information Science and Engineering, Northeastern University, Shenyang, Liaoning 110819, China
    • show less
    AbstractGet PDF(in Chinese)
    Figures&Tables (15)
    Equations (0)
    References (26)
    Cited By (4)
    Tools
    Paper Information
    Topics
    Figures & Tables(15)
    FCN structural diagram

    Fig. 1. FCN structural diagram

    Download full size
    ResNet basic structural unit

    Fig. 2. ResNet basic structural unit

    Download full size
    Diagram of ResNet18 structure

    Fig. 3.

    Diagram of ResNet18 structure

    Download full size
    Main structure of network

    Fig. 4. Main structure of network

    Download full size
    Diagram of N×N channel structure

    Fig. 5. Diagram of N×N channel structure

    Download full size
    Two-channel feature fusion with different scales

    Fig. 6. Two-channel feature fusion with different scales

    Download full size
    Downsampling and upsampling of location index max pool

    Fig. 7. Downsampling and upsampling of location index max pool

    Download full size
    Partial images of data sets and corresponding visual labels

    Fig. 8. Partial images of data sets and corresponding visual labels

    Download full size
    Partial data set images and their visual mark display after cutting

    Fig. 9. Partial data set images and their visual mark display after cutting

    Download full size
    Partial original images and display of flipped and rotated images

    Fig. 10. Partial original images and display of flipped and rotated images

    Download full size
    Visual classification results of remote sensing images by different classification algorithms

    Fig. 11. Visual classification results of remote sensing images by different classification algorithms

    Download full size
    Visual classification results of channel 1, channel 1+2, and overall structure

    Fig. 12. Visual classification results of channel 1, channel 1+2, and overall structure

    Download full size

    • Table 1. Obfuscation matrix of classification results of proposed algorithmpixel

      View table

      Table 1. Obfuscation matrix of classification results of proposed algorithmpixel

      CategoryVegetationBuildingWaterRoadOthers
      Vegetation9225390393999718781875047281427
      Building7837261250580050563126407290737
      Water18293690152069612676129298
      Road201965049050493256576176714
      Others834751446908348417382672437

    • Table 2. Classification accuracy and RKappa of different algorithms

      View table

      Table 2. Classification accuracy and RKappa of different algorithms

      AlgorithmClassification accuracy /%RKappa
      VegetationBuildingWaterRoadOthersOA
      FCN-8s71.8871.4370.0969.9771.2371.300.6603
      Unet80.9280.2780.1079.8980.5280.440.7522
      SegNet83.6683.1882.6482.5283.0783.210.7810
      Channel 181.1681.2381.4780.9781.3281.200.7623
      Channel 1+289.5689.7889.5889.1689.7689.630.8406
      Ours90.7790.8290.1890.3090.5790.680.8595

    • Table 3. Convolution kernel parameters, single forward propagation time, and training time of different algorithms

      View table

      Table 3. Convolution kernel parameters, single forward propagation time, and training time of different algorithms

      AlgorithmTotal number of parameters /106Forward time /msTrain time /h
      FCN-8s134.32215.56
      Unet23.6564.72
      SegNet29.4784.88
      Channel 115.3504.55
      Channel 1+230.51085.06
      Ours30.51162.90
    Tools

    Get Citation

    Copy Citation Text

    Ende Wang, Kai Qi, Xuepeng Li, Liangyu Peng. Semantic Segmentation of Remote Sensing Image Based on Neural Network[J]. Acta Optica Sinica, 2019, 39(12): 1210001

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Set citation alerts for article
    Save article for my favorites
    Paper Information

    Category: Image Processing

    Received: Jul. 9, 2019

    Accepted: Aug. 19, 2019

    Published Online: Dec. 6, 2019

    The Author Email: Qi Kai (qiqikai123456@163.com)

    DOI:10.3788/AOS201939.1210001

    Topics

    laser devices and laser physics
    Lasers and Laser Optics
    Laser physics
    laser manufacturing
    Instrumentation, Measurement and Metrology