This paper proposed a novel defect detection system based on deep learning and focused on a practical industrial application: steel plate defect inspection and employs a baseline convolution neural network to generate feature maps at each stage, and the proposed multilevel feature fusion network (MFN) combines multiple hierarchical features into one feature, which can include more location details of defects.
Abstract:
A complete defect detection task aims to achieve the specific class and precise location of each defect in an image, which makes it still challenging for applying this task in practice. The defect detection is a composite task of classification and location, leading to related methods is often hard to take into account the accuracy of both. The implementation of defect detection depends on a special detection data set that contains expensive manual annotations. In this paper, we proposed a novel defect detection system based on deep learning and focused on a practical industrial application: steel plate defect inspection. In order to achieve strong classification ability, this system employs a baseline convolution neural network (CNN) to generate feature maps at each stage, and then the proposed multilevel feature fusion network (MFN) combines multiple hierarchical features into one feature, which can include more location details of defects. Based on these multilevel features, a region proposal network (RPN) is adopted to generate regions of interest (ROIs). For each ROI, a detector, consisting of a classifier and a bounding box regressor, produces the final detection results. Finally, we set up a defect detection data set NEU-DET for training and evaluating our method. On the NEU-DET, our method achieves 74.8/82.3 mAP with baseline networks ResNet34/50 by using 300 proposals. In addition, by using only 50 proposals, our method can detect at 20 ft/s on a single GPU and reach 92% of the above performance, hence the potential for real-time detection.
TL;DR: Comprehensive and up-to-date, this book includes essential topics that either reflect practical significance or are of theoretical importance and describes numerous important application areas such as image based rendering and digital libraries.
TL;DR: A pyramid feature fusion and global context attention network for pixel-wise detection of surface defect, called PGA-Net, which outperforms the state-of-the-art methods on mean intersection of union and mean pixel accuracy.
TL;DR: In this paper, a survey of state-of-the-art deep learning methods for defect detection is presented, focusing on three aspects, namely method and experimental results, and the core ideas and codes of studies related to high precision, high positioning, rapid detection, small object, complex background, occluded object detection and object association.
TL;DR: Compared with the existing saliency detection methods, the deeply supervised EDRNet can accurately segment the complete defect objects with well-defined boundary and effectively filter out irrelevant background noise.
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
TL;DR: The state-of-the-art performance of CNNs was achieved by Deep Convolutional Neural Networks (DCNNs) as discussed by the authors, which consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax.
TL;DR: In this paper, the authors investigated the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting and showed that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 layers.
TL;DR: A new database called “ImageNet” is introduced, a large-scale ontology of images built upon the backbone of the WordNet structure, much larger in scale and diversity and much more accurate than the current image datasets.
TL;DR: Deep learning is making major advances in solving problems that have resisted the best attempts of the artificial intelligence community for many years, and will have many more successes in the near future because it requires very little engineering by hand and can easily take advantage of increases in the amount of available computation and data.
Q1. What contributions have the authors mentioned in the paper "An end-to-end steel surface defect detection approach via fusing multiple hierarchical features" ?
In this paper, the authors proposed a novel defect detection system based on deep learning and focused on a practical industrial application: steel plate defect inspection. In order to achieve strong classification ability, this system employs a baseline convolution neural network ( CNN ) to generate feature maps at each stage, and then the proposed multilevel feature fusion network ( MFN ) combines multiple hierarchical features into one feature, which can include more location details of defects. In addition, by using only 50 proposals, their method can detect at 20 ft/s on a single GPU and reach 92 % of the above performance, hence the potential for real-time detection.
Q2. What is the importance of depth of networks?
The early successful networks are based on the sequential pipeline architecture [25], which establish the basic structure of CNN and prove the importance of depth of networks.
Q3. What is the purpose of the feature?
In the feature, the authors will focus on two directions as follows: the one is data augmentation technology due to the expensive manual annotations in detection data sets.
Q4. What types of defects are found in hot-rolled steel plates?
There are six types of defects from hot-rolled steel plates, including crazing, inclusion, patches, pitted surface, rolled-in scales, and scratches.
Q5. What is the way to solve the defect classification task?
To solve it, the simple and direct way is to perform defect localization before defect classification making the inspection task classify on regions of defects instead of a whole defect image, which is the defect detection task.
Q6. What is the importance of pretraining on the ImageNet data set?
As the authors know that pretraining on the ImageNet data set is important to achieve competitive performance, and then this pretrained model can be fine-tuned on a relatively small defect data set.
Q7. How many images are used for fine-tuning the network?
The training set containing 1260 images used for fine-tuning the network introduced in Section IV-B, and the test set containing 540 images.
Q8. What is the way to improve the recall of a defect detection method?
Increasing the number of proposals can get a promising recall, but this will greatly increase the runtime of the detection [38], and what is worse, low-quality proposals would be involved in the process of detection, leading to failure of defect detection in some cases.
Q9. How do the authors fine tune DDN using top-300 region proposals?
In the following, the authors fine-tune DDN using top-300 region proposals owing to the extracted quality region proposals, but reduce this number to accelerate the detection speed without harming accuracy at test-time.