Aiming at the problems of low efficiency, high false detection rate, and poor real-time performance of current industrial defect detection methods, this paper proposes an industrial defect detection method based on an expanded perceptual field and feature fusion for practical industrial applications. First, to improve the real-time performance of the network, the original network structure is enhanced by using depth-separable convolution to reduce the computation while ensuring the detection accuracy, and the critical information extraction from the feature map is enhanced by using MECA (More Efficient Channel Attention) attention to the detection network. To reduce the loss of small target detail information caused by the pooling operation, the ASPF (Atrous Spatial Pyramid Fast) module is constructed using dilate convolution with different void rates to extract more contextual information. Secondly, a new feature fusion method is proposed to fuse more detailed information by introducing a shallower feature map and using a dense multiscale weighting method to improve detection accuracy. Finally, in the model optimization process, the K-means++ algorithm is used to reconstruct the prediction frame to speed up the model’s convergence and verify the effectiveness of the combination of the Mish activation function and the SIoU loss function. The NEU-DET steel dataset and PCB dataset is used to test the effectiveness of the proposed model, and compared to the original YOLOv5s, our method in terms of mAP metrics by 6.5% and 1.4%, and in F1 by 5.74% and 1.33%, enabling fast detection of industrial surface defects to meet the needs of real industry.
Background
The three-dimensional (3D) printing technology has remarkable potential as an auxiliary tool for representing anatomical structures, facilitating diagnosis and therapy, and enhancing training and teaching in the medical field. As the most available diagnostic tool and it is routinely used as the first approach in diagnosis of the uterine anomalies, 3D transvaginal ultrasonography (3D-TVS) has been proposed as non-invasive “gold standard” approach for these malformations due to high diagnostic accuracy. Despite holding promise of manufacturing 3D printed models based on 3D-TVS data, relevant reports about 3D-TVS derived gynecological 3D printing haven’t been reported to the best of our knowledge. We found an opportunity to explore the feasibility of building 3D printed models for the abnormal uterus based on the data acquired by 3D-TVS.
Methods
The women suspected with congenital uterine anomalies (CUAs) were enrolled in the study. The diagnose of CUAs were made by 3D-TVS scanning and further confirmed under the hysteroscopy examination. One volunteer with normal uterus was enrolled as control. All subjects underwent 3D-TVS scanning for 3D printing data collection. Acquired images were stored and extracted as DICOM files, then processed by professional software to portray and model the boundary of the uterine inner and outer walls separately. After the computer 3D models were constructed, the data were saved and output as STL files for further surface restoration and smoothing. The colors of endometrium and uterine body were specified, respectively, in the print preview mode. Then the uncured photosensitive resin was cleaned and polished to obtain a smooth and transparent solid model after printed models were cooled down.
Results
3D printing models of normal uterus, incomplete septate uterus, complete septate uterus, uterus didelphys and unicornuate uterus were produced on ultrasonographic data of 3D-TVS.
Conclusions
Our research and practice made the first try in modeling CUAs successfully based on ultrasonographic data entirely, verifying that it’s a feasible way to build 3D printed models of high-quality through 3D-TVS scanning.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.