Search

• A method of removing the interference on the boundary of the printing surface using the height difference constraint is proposed. • The point cloud rows containing defects extracted using the extreme difference of height values are used as the object of study in this paper. • Divide the point cloud rows into segments and recombine them according to different boundaries. • The characteristic parameter matrix of defects is constructed and divided into three main types. Defect detection is a crucial part of the additive manufacturing process. Based on 3D laser point cloud, this paper takes the surface of AM process as the research object for defect detection, including surface point cloud pretreatment and internal defect detection. Aiming at a kind of interference existing on the boundary in the process of AM, we use continuous constraint and improved slope method to deal with it, and discuss two cases of continuous surface and non-continuous surface. The small threshold based pass-through filtering algorithm extracts the external points of the defect boundary by using the fluctuation of the height value, and classifies the defect point cloud by classifying, combining and matching the external points. Finally, experimental verification results show that the accuracy of our proposed surface defect detection method can reach 92.3%. This method can effectively extract relatively complete defects. It provides important technical support for surface quality inspection in industrial automation. [ABSTRACT FROM AUTHOR]

The machined surface in the industrial cutting environment is often covered by some interference factors, such as chips or coolant residuals, which make the existing methods have large detection errors. Therefore, the interference factors in the cutting environment seriously affect the sustainability detection and evaluation of the processed surface quality. This paper discusses the relationship between interference factors and roughness to study the sustainability detection of roughness. Firstly, a multidimensional feature parameter matrix with strong correlation with surface roughness is extracted and constructed based on the gray level co-occurrence matrix. On this basis, the adverse effects of interference factors (chips) on image feature parameters are quantitatively analyzed. According to the relationship between chip area and error change rate of feature parameters, an error correction model is constructed to optimize the feature parameters that change due to interference factors. The error correction model greatly reduces the negative influence of chip interference. This is the core of this paper. Further, the BP neural network model and support vector machine (SVM) model are used to predict the surface roughness with the optimized multi-dimensional feature parameters matrix as input, respectively. The above process realizes the sustainable detection of machined surface roughness. At the same time, in order to facilitate industrial applications, this paper uses LabVIEW software and MATLAB software to package the above research into a software system. Finally, the practicability and effectiveness of the sustainable detection research are verified by practical application in the industrial scene. This study promotes the sustainable development between product safety and environmental impact in the industrial manufacturing process. [ABSTRACT FROM AUTHOR]

The inherent brittleness of 3D-printed ceramic parts makes the surface more prone to produce defects. In the 3D-printing environment, the defect regions in the surface are easily covered by interference factors. This brings great difficulties to surface defect detection. Based on this, this paper proposes an anti-interference detection method for surface defects of ceramic parts based on deep learning. This anti-interference detection method is mainly divided into three stages: interference factors identification, interference factors repair, defect detection. Interfering factors in the surface are located and identified through the built Multimodal Feature Layer Fusion-PSP network (MFLF-PSP net) model. MFLF-PSP net's mAP for interference factor identification is up to 95.67%. Then, based on the results of the interference identification, the proposed Parallel Spatial-Channel Attention Mechanism (PSCAM)-RFR net model is used to perform pixel filling and repair in the regions where the interference factor is located. This solves the difficult problem of defect detection caused by interference factors. On this basis, the constructed Inception-SSD network model is used to perform effective defect detection on ceramic surface images. The mAP of the model for the detection of crack defects and bulge defects is 96.34% and 94.92%, respectively. Through the close cooperation of the above three stages, the problem of defect detection caused by interference factors, such as misjudgment and defect region segmentation, has been solved. It provides technical support for automatic detection of ceramic surface quality during 3D-printing. [ABSTRACT FROM AUTHOR]

3D printed ceramic parts often have defects due to their inherent brittleness. These defects are of various scale sizes. Detecting these defects, especially with multiple sizes, is a challenging task in the field of detection. Furthermore, there is a lack of surface defect real-time detection systems suitable for industrial applications in this field. Based on this, this paper makes two contributions: the design and development of a real-time defect detection system, and the proposal of a multiscale defect detection method for 3D printed ceramic surfaces within this system. Firstly, the paper establishes the overall structure of the real-time detection system for surface defects on 3D printed ceramic components, and describes the hardware and software components of the system. On this basis, a dataset of ceramic surface defect images is collected and constructed. Then, experimental analyses point out the shortcomings of the You Only Look Once version-5 (YOLOv5) model for multiscale defect detection. To address the shortcomings, the YOLOv5 model is optimized from three aspects, resulting in the Deep separable convolution + residual network-SKNetwork-Efficient Channel Attention Network-YOLOv5 (DepRes-SK-ECA-YOLOv5) model for multiscale defect detection. This model improves the ability to extract and fuse features of defects at different scales. The experimental results show that the DepRes-SK-ECA-YOLOv5 model can achieve 93.5 %, 91.6 %, 94.3 %, and 0.198 s for Precision, Recall, mAP, and Speed for the test set, respectively. Finally, the paper designs the workflow for the system software. The system hardware and system software are integrated to form a real-time detection system. The performance of the detection system is verified through experiments. [ABSTRACT FROM AUTHOR]

On account of the complex lighting environmental impact of the machine tool, the tool body and wear boundary cannot be accurately extracted, which brings adverse effects to the wear detection. At the same time, most of the research on the side edge wear detection of end mill has focused on local wear, lacking the judgment and analysis of the overall side edge wear. In view of this situation, a "holistic to local" tool wear detection method based on the reconstruction idea was proposed. Firstly, the wear condition of the tool flank was identified as a whole, and then, the wear area affecting the remaining useful life of the tool was detected. The method mainly consists of three steps: the extraction of tool body, the reconstruction of tool wear edge, and tool wear detection. Aiming at the complex background of the tool image, a semantic segmentation algorithm based on U-net neural network was applied to separate foreground and background and extract the tool body. Secondly, based on pixel accuracy, the tool edge was extracted and the tool wear boundary was reconstructed. Finally, the range of the wear boundary coordinates was obtained, the number of pixels occupied by the maximum wear amount was gained, and the maximum wear of the tool was calculated according to the pixel length equivalent. Comparing the wear amount detected by the algorithm with the actual measured value, the results show that the method has high detection accuracy and the average error is controlled within 5%, which proves the effectiveness of the method and realizes the high-precision detection of the wear state of the tool during machining. [ABSTRACT FROM AUTHOR]