Prereflow automatic optical inspection (AOI) has been widely used to ensure product quality in surface mount technology (SMT). When confronted with a complex industrial environment, traditional hand-designed visual inspection algorithms may lack robustness and generalizability. In this article, PCBNet, a convolutional neural network (CNN) method that combines data preprocessing, detection network, and visualization, is proposed to localize electronic components and recognize defects. In the data preprocessing stage, raw images are segmented into several regions of interest (ROIs). The ROI patches are inspected by a CNN-based detection system, which is capable of classifying defects and positioning components. After inspection, the reporting system visualizes the results via the human–computer interface. In comparative studies, the effectiveness of the proposed PCBNet was validated on a large-scale PCB component defect dataset. The PCBNet backbone outperforms other well-known lightweight CNN backbones in terms of accuracy and latency on $4\times $ ARM Cortex A72 CPU @ 1.5 GHz. Compared to other learning-based methods on the small-scale benchmark dataset, the PCBNet also achieves the best balance between inference speed and accuracy. In addition, extensive experiments demonstrate the superior efficiency of PCBNet in comparison to some famous traditional object detectors and novel oriented object detection algorithms.

Machining feature recognition is a research hotspot in recent years. A point cloud is a geometry data representation format of three-dimensional (3D) models. The use of point cloud-based convolutional neural networks (CNNs) for machining feature recognition has received increasing research attention. However, these point cloud-based networks usually have large complexity size and training time. In this paper, a selective downsampling-based point neural network for machining feature recognition is proposed. Firstly, a machining feature dataset called MFDataset is constructed and contains 33 feature types. Secondly, a selective downsampling algorithm of the input points is presented, which drops out unimportant points while keeping the important ones. In single-machining feature recognition, MFPointNet is proposed by utilizing the selective downsampling of the input points. In multi-machining feature recognition, the segmentation part of the MFPointNet is adopted with the selective downsampling algorithm to segment and recognize multiple features. Compared with other point cloud-based networks, experimental results show that MFPointNet reduces the computational complexity without losing the recognition accuracy basically. MFPointNet is more robust to model complexity when more machining feature points are input to the network. Moreover, several intersecting feature models validate the segmentation performance of MFPointNet.

Automated machining feature recognition is an essential component linking computer-aided design (CAD) and computer-aided process planning (CAPP). Deep learning (DL) has recently emerged as a promising method to improve machining feature recognition. However, training DL-based recognition models typically require annotating large amounts of data, which is time-consuming and labor-intensive for researchers. Additionally, DL models struggle to achieve satisfactory results when presented with small labeled datasets. Furthermore, existing DL-based approaches require significant memory and processing time, thus hindering their real-world application. To address these challenges, this paper presents a semi-supervised learning framework that leverages both labeled and unlabeled data to learn meaningful visual representations. Specifically, self-supervised learning is utilized to extract prior knowledge from a large dataset without annotations, which is then transferred to improve downstream feature recognition tasks. Furthermore, we apply lightweight network techniques to two established feature recognizers, FeatureNet and MsvNet, to develop reduced-memory, computationally efficient models termed FeatureNetLite and MsvNetLite, respectively. To validate the effectiveness of the proposed approaches, we conducted comparative studies on the FeatureNet dataset. With only one training sample per class, MsvNetLite outperformed MsvNet by about 19%, whereas FeatureNetLite outperformed FeatureNet by approximately 20% in machining feature classification. On a common X86 CPU, MsvNetLite gained 6.68× improvement in speed over MsvNet, and FeatureNetLite was 2.49× faster than FeatureNet. The proposed semi-supervised learning framework shows a significant improvement in machining feature recognition on small labeled data while achieving the optimal balance between recognition accuracy and inference speed compared to other DL-based approaches.