Abstract:Deep-learning-based detection methods have been widely applied to industrial defect inspection. However, directly using vanilla detection methods fails to achieve satisfying performance due to the lack of identifiable features. In this article, a novel attention-based multiscale feature fusion module (AMFF) is proposed, aiming to enhance defect features and improve defect identification by leveraging attention mechanism in the feature fusion. AMFF includes self-enhanced attention module (SEAM) and cross-enhanced attention module (CEAM). SEAM is performed on a single feature map, which first adopts multiple dilation convolutions to enrich contextual information without compromising resolution and then utilizes the intralayer attention on the current feature map. CEAM takes both the current feature map and the adjacent feature map as input to perform cross-layer attention. The adjacent feature map is modulated with the guidance of the current feature map, which is then combined with the current feature map and the output of SEAM for final prediction. AMFF is utilized in current feature fusion networks, e.g., feature pyramid network (FPN) and path aggregation FPN (PAFPN), and is further integrated into prevalent detectors to guide them to pay more attention to defects rather than the background. Extensive experiments are conducted on two real industrial datasets released by Tianchi platform, i.e., fabric and aluminum defect datasets. For each dataset, 500 images are randomly selected for test and the rest for training. The proposed AMFF is demonstrated to significantly boost defect detection accuracy with acceptable computational cost, and the real-time performance could fully satisfy practical requirements.

Study on Multi-Scale AMSS Operator for Fabric Flaw Detection

Adaptive Multi-Scale Amss Operator for Quality Detection of Silks

Study on Flaw Labeling and Edge Tracking in Ultrasonic C-Scan Image

Multivariate image analysis in gaussian multi-scale space for defect detection

A flaw echo detection method based on S-transformation and time-frequency spectrum spatial filtering

Flaw Edge-detection and Feature Parameter Construction in Ultrasonic C-scan Image

A multivariable statistical process monitoring method based on multiscale analysis

Application Study on Fuzzy Multiscale Edge Detection Algorithm

Fabric Defect Detection Based on Multi-Scale Wavelet Transform and Gaussian Mixture Model Method

Textural image segmentation with multi-scale wavelet analysis based on feature learning

Visualizing Defects of Irregular Weld Seams Based on MultiMode TFM Detection

Fabric Defect Detection Based on Wavelet Decomposition with One Resolution Level

Fabric Edge Cutting Algorithm Based on Multiscale Feature Fusion

Automatic Detection on Texture Defects of Fabric Surface

An Investigation Of Data Fusion For Defects Inspection In Additive Manufacturing

Weld Slim Line Defects Extraction Based on Adaptive Local Threshold and Modified Hough Transform

Image Denoising Method on Surface of Steel Strip Based on Partial Differential Equations

Multi-scale Coefficients Fusion Strategy for Enhancement of SAM Image in Solder Joints Detection

Attention-Based Multiscale Feature Fusion for Efficient Surface Defect Detection

Frequency-Diverse Geometric-Mean And Arithmetic-Mean Filtering For Ultrasonic Flaw Detection

An Infrared Thermal Image Processing Framework Based on Superpixel Algorithm to Detect Cracks on Metal Surface