Skip to content

ZeaLousGeek/Ship-Image-Classification

Repository files navigation

Including files

main.py: includes data loader, moodel loader, train, and test and draw figure

  • model include
    • Alexnet
    • VGGNet11
    • ResNet18
    • DenseNet
  • draw figure
    • Loss
    • accuracy top1, top5
    • classification confusion data ##num_pre: includes
  • accuracy_topk: computes the precision of top1 and top5

draw_figure.py:includes

  • plot_with_labels:
  • images_example:

self_dataset_loader: includes

  • class TorchSelfDataset: dataset loader

self_image_processing:include

  • show_image:
    • show RGB image in matplotlib
  • cv_show_image:
    • show RGB image in openCV
  • read_image:
    • read image in openCV
    • resize image
    • normalization of image
  • fast_read_image_roi:
    • read image in openCV
    • fixed resize image, (IMREAD_REDUCED mode, 1/2, 1/4, 1/8)
    • normalization of image
  • resize_image:
    • resize image
  • scale_image:
    • resize image
  • get_rect_image:
    • get a rectangle image
  • scale_rect:
    • scale original image and rectangle
  • show_image_rect:
    • show rectangle, not cat rectangle image
    • no return parameters
  • rgb_to_gray:
    • image from RGB to GRAY
  • save_image:
    • save image
  • combime_save_image:
    • save image and modify file name

About

This is an alogrithm of CNN for ship classification

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages