It is difficult to detect defects such as shallow dents and rust on rolling surfaces by using the traditional vision-based surface inspection method (line light source and line array camera mode) which has a low sensitivity to these defects. This paper presents a method that introduces the fringe projection technique for traditional visual inspection devices to overcome the limitations of the traditional methods and uses deep-learning techniques for detecting defects such as cuts, abrasions, dents, and rust on the rolling surfaces of drum-shaped rollers. A new artificial-intelligence-based labeling method, namely, the Padua Incremental Mask Labeling Method, has been introduced for accelerating the calibration process used for defect detection, and based on a one-stage architecture, the You-Only-Look-Once-OurNet (YOLO-OurNet) deep-learning network has been designed for detecting the defects on the rolling surfaces of drum-shaped rollers. From the results of the experimental tests, the time required for detecting a defect has been found to be 0.024s, an accuracy rate of up to 99.2%, and the value of object detection evaluation index F1 of up to 0.988. Our method outperforms the related method on the domain of rolling surface defect detection.

For the current problems of low accuracy and poor reliability of defect detection for bearing roller end surfaces in industrial production, this paper proposes a bearing roller end surface defect detection algorithm based on improved YOLOv5n and the fusion of gamma-corrected maps and curvature maps. First, this paper uses photometric stereo vision to reconstruct the three-dimensional shape of the surface and proposes an improved Frankot–Chellappa integration algorithm to solve the problem of reconstructing surface deformation. Secondly, the DenseFuse network is used to fuse gamma-corrected maps and curvature maps to generate an image dataset that combines the strengths of both images to enhance defect features and improve the precision of target detection. Finally, the improved target detection model YOLOv5n is proposed to detect defects in the end surfaces of bearing rollers. The experimental results show that by using fused images for training, detection models with higher mean average precision(mAP) than traditional images can be obtained, and the improved YOLOv5n algorithm maintains the high real-time performance of the original algorithm while the mAP0.5 and mAP0.5:0.95 of improved YOLOv5n are 98.6% and 87.4%, respectively, which are respectively 0.9% and 2.8% higher than YOLOv5n.

Object detection is widely used in the industry for visual inspection of defects. But it fails to fully reflect the defects on the smooth surface due to the difficulty in illuminating, and the high performance of defect detection tends to be a challenge. In this paper, we proposed a novel defect detection method for laptop panels. On the one hand, we utilized the wrapped phase map obtained by phase-shift reflection fringes for defect detection and used the DenseFuse network to fuse the wrapped phase map with the fringe modulation map to optimize the dataset. On the other hand, we proposed an improved deep learning-based network Vovecwnet. The experiments indicated that our image-based multi-information fusion system could achieve excellent performance in defect detection on the smooth surface (e.g., laptop panels surface), with high accuracy of 97.5% and a few reference time 0.0762s on GTX 2080ti GPU. Thus, our proposed method outperformed and tackled the challenges encountered in the defect detection of smooth surfaces.

To tackle the challenges in the current detection of patterned fabric defects, such as unbalanced and limited data volume, incomplete extraction of defect features by the network, aliasing effect in feature fusion by the pyramid network, and low accuracy in identifying small targets, this paper proposes a novel single-stage network named ASC-you only look once (YOLO). Firstly, the CoT-CSP module is introduced into the Backbone network of the YOLOv5s to model the global contextual relationships of defective features. This module compensates for the limitations of the original Cross-Stage Partial Network, which can only extract local feature information. It effectively enhances the comprehensive extraction of various irregular defect features in fabrics and reduces redundant interactions with texture background noise. Subsequently, three types of AttentionC3 modules are proposed and integrated into the Neck enables the model to focus on valuable details related to defect features through weight allocation, significantly enhancing the capability to identify small defects. Finally, an adaptive spatial feature fusion network is designed to optimize the pyramid network by redistributing weights across different feature layers for efficient utilization of multi-level feature information. Additionally, a random erasing approach is proposed to individually enhance the defect images of specific types with a limited number and adjust the corresponding labels simultaneously. This process aims to improve the robustness of the ASC-YOLO network against data noise. Experimental results demonstrate that ASC-YOLO achieves 95.8% accuracy, with mean average precision@0.5 reaching 83.1%. Meanwhile, the experiments also verify its superiority over other fabric detection methods and its ability to effectively improve the accuracy of fabric defect detection.

Timely detection and management of daylily diseases are crucial to prevent yield reduction. However, detection models often struggle with handling the interference of complex backgrounds, leading to low accuracy, especially in detecting small targets. To address this problem, we propose DaylilyNet, an object detection algorithm that uses multi-task learning to optimize the detection process. By incorporating a semantic segmentation loss function, the model focuses its attention on diseased leaf regions, while a spatial global feature extractor enhances interactions between leaf and background areas. Additionally, a feature alignment module improves localization accuracy by mitigating feature misalignment. To investigate the impact of information loss on model detection performance, we created two datasets. One dataset, referred to as the ‘sliding window dataset’, was obtained by splitting the original-resolution images using a sliding window. The other dataset, known as the ‘non-sliding window dataset’, was obtained by downsampling the images. Experimental results in the ‘sliding window dataset’ and the ‘non-sliding window dataset’ demonstrate that DaylilyNet outperforms YOLOv5-L in mAP@0.5 by 5.2% and 4.0%, while reducing parameters and time cost. Compared to other models, our model maintains an advantage even in scenarios where there is missing information in the training dataset.