• Journal of Infrared and Millimeter Waves
  • Vol. 39, Issue 5, 650 (2020)
Li-Li DONG, Tong ZHANG*, Dong-Dong MA, and Wen-Hai XU
Author Affiliations
  • School of Information Science and Technology, Dalian Maritime University, Dalian116026, China
  • show less
    DOI: 10.11972/j.issn.1001-9014.2020.05.016 Cite this Article
    Li-Li DONG, Tong ZHANG, Dong-Dong MA, Wen-Hai XU. Maritime background infrared imagery classification based on histogram of oriented gradient and local contrast features[J]. Journal of Infrared and Millimeter Waves, 2020, 39(5): 650 Copy Citation Text show less
    Sea infrared images
    Fig. 1. Sea infrared images
    Three-dimensional map of sea infrared images
    Fig. 2. Three-dimensional map of sea infrared images
    SSLE
    Fig. 3. SSLE
    The process of extracting features
    Fig. 4. The process of extracting features
    Fused features
    Fig. 5. Fused features
    Two similar gray distribution of different background images
    Fig. 6. Two similar gray distribution of different background images
    Gray distribution of base layer and detail layer of different background images
    Fig. 7. Gray distribution of base layer and detail layer of different background images
    Local contrast calculation diagram
    Fig. 8. Local contrast calculation diagram
    16*16 matrix
    Fig. 9. 16*16 matrix
    Local contrast calculation diagram based on different sea backgrounds:
    Fig. 10. Local contrast calculation diagram based on different sea backgrounds:
    Accuracy and time cost of different classification algorithms
    Fig. 11. Accuracy and time cost of different classification algorithms
    Cell size8*816*1632*32
    Feature vector dimension3459681001764
    Time cost(s)48.11915.65312.515
    Table 1. Comparison of different cell sizes
    Number of misjudged imagesClassification error rate
    Total images93.6%
    Classification of different types of background images20.8%
    Classification accuracy of multiple types of backgrounds in the same image72.8%
    Table 2. Classification accuracy of different background images
    ImageExpected classificationActual classification
    SFESSLE
    Table 3. A misclassified image
    Li-Li DONG, Tong ZHANG, Dong-Dong MA, Wen-Hai XU. Maritime background infrared imagery classification based on histogram of oriented gradient and local contrast features[J]. Journal of Infrared and Millimeter Waves, 2020, 39(5): 650
    Download Citation