Method for extracting container loading bridge based on remote sensing interpretation analysis technology

Method for extracting container loading bridge based on remote sensing interpretation analysis technology

  • CN 108,257,143 B
  • Filed: 12/12/2017
  • Issued: 09/15/2020
  • Est. Priority Date: 12/12/2017
  • Status: Active Grant
First Claim
Patent Images

1. A method for extracting a container loading bridge based on a remote sensing interpretation analysis technology is characterized by comprising the following steps:

  • s100, performing target detection based on texture features, constructing a background model through unchanged LBP texture gray, extracting a foreground probability map from the background model, performing difference to obtain an initial target model, and describing an image by replacing specific data with a sign function of interpolation of a central pixel and a neighborhood pixel in the difference process;

    s200, image segmentation, namely firstly determining segmentation parameters on an initial target model, homogenizing the segmentation parameters through a homogenizing factor according to a remote sensing image, then dividing the image into a plurality of non-overlapping regions according to different characteristics, wherein the characteristics comprise gray scale, space texture and geometric shape, and based on a segmentation algorithm, performing region combination on two factors of spectrum and space information during specific segmentation, and performing feature comparison according to the same comparison principle during comparison;

    s300, topology and extraction of the remote sensing image are carried out, topology is carried out in sequence from the segmented image area, in topology calculation, difference assignment needs to be carried out on the remote sensing image in sequence through principal component analysis, a plurality of principal component images are obtained in sequence during principal component analysis, intersection and redundancy of information provided by wave bands are reduced, weight of the principal wave bands is obtained in the calculation process, shape features in the remote sensing image are extracted, normalization processing is carried out to obtain a shape template, and the shape template and a target shape do not influence the calculation of shape distance after the normalization processing after the shape features are extracted;

    s400, compensating the shape template according to the texture characteristics to identify the container loading and unloading bridge, and compensating the texture characteristics changed by LBP on the basis of the shape template one by one to accurately identify the container loading and unloading bridge.

View all claims
    ×
    ×

    Thank you for your feedback

    ×
    ×