Xdefects

Automatic defect recognition in X-ray testing using computer vision

Requirements

This code needs the following toolboxes:

• MatConvNet
• VLFeat
• SPAMS
• Neural Networks
• Computer Vision
• LIBSVM
• Balu
• Experimental

Original images are from GDXray

http://dmery.ing.puc.cl/index.php/material/gdxray/

0. Dataset:

   see imdb.mat and imdb_readme.txt

1. Code for classical feature extraction and classification:

   see wacv_demo.m and wacv_examples.m in this example the following pretrained nets must be included in folder nets: imagenet-caffe-alex.mat imagenet-vgg-f.mat imagenet-vgg-verydeep-16.mat imagenet-googlenet-dag.mat imagenet-vgg-m-2048.mat imagenet-vgg-verydeep-19.mat please see vlfeat.org for downloading these files.

2. Code fox Xnet:

   see folder xnet and use code xnet_main.m for training and testing.

3. Sliding windows for detection in one X-ray image:

   see wacv_sliwin.m

Reference

Mery, D.; Arteta, C.: Automatic Defect Recognition in X-ray Testing using Computer Vision. In 2017 IEEE Winter Conference on Applications of Computer Vision, WACV2017.

(c) 2017 - Domingo Mery and Carlos Artera