scholarly journals Real-time detection of particleboard surface defects based on improved YOLOV5 target detection

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Ziyu Zhao ◽  
Xiaoxia Yang ◽  
Yucheng Zhou ◽  
Qinqian Sun ◽  
Zhedong Ge ◽  
...  
Keyword(s):  
Real Time ◽  
Gamma Ray ◽  
Surface Defects ◽  
Detection Algorithm ◽  
Single Image ◽  
Transform Method ◽  
Ray Transform ◽  
Detection Technology ◽  
Embedded Model ◽  
Surface Defect Detection

AbstractParticleboard surface defect detection technology is of great significance to the automation of particleboard detection, but the current detection technology has disadvantages such as low accuracy and poor real-time performance. Therefore, this paper proposes an improved lightweight detection method of You Only Live Once v5 (YOLOv5), namely PB-YOLOv5 (Particle Board-YOLOv5). Firstly, the gamma-ray transform method and the image difference method are combined to deal with the uneven illumination of the acquired images, so that the uneven illumination is well corrected. Secondly, Ghost Bottleneck lightweight deep convolution module is added to Backbone module and Neck module of YOLOv5 detection algorithm to reduce model volume. Thirdly, the SELayer module of attention mechanism is added into Backbone module. Finally, replace Conv in Neck module with depthwise convolution (DWConv) to compress network parameters. The experimental results show that the PB-YOLOv5 model proposed in this paper can accurately identify five types of defects on the particleboard surface: Bigshavings, SandLeakage, GlueSpot, Soft and OliPollution, and meet the real-time requirements. Specifically, recall, F1 score, [email protected], [email protected]:.95 values of pB-Yolov5s model were 91.22%, 94.5%, 92.1%, 92.8% and 67.8%, respectively. The results of Soft defects were 92.8%, 97.9%, 95.3%, 99.0% and 81.7%, respectively. The detection of single image time of the model is only 0.031 s, and the weight size of the model is only 5.4 MB. Compared with the original YOLOv5s, YOLOv4, YOLOv3 and Faster RCNN, the PB-Yolov5s model has the fastest Detection of single image time. The Detection of single image time was accelerated by 34.0%, 55.1%, 64.4% and 87.9%, and the weight size of the model is compressed by 62.5%, 97.7%, 97.8% and 98.9%, respectively. The mAP value increased by 2.3%, 4.69%, 7.98% and 13.05%, respectively. The results show that the PB-YOLOV5 model proposed in this paper can realize the rapid and accurate detection of particleboard surface defects, and fully meet the requirements of lightweight embedded model.

Lightweight detection algorithm for fine-grained surface defects of aerospace seal rings

2021 ◽  
pp. 1-18
Author(s):  
Hui Liu ◽  
Boxia He ◽  
Yong He ◽  
Xiaotian Tao
Keyword(s):  
Feature Extraction ◽  
Defect Detection ◽  
Surface Defect ◽  
Surface Defects ◽  
Detection Algorithm ◽  
Seal Ring ◽  
Fine Grained ◽  
Aerospace Applications ◽  
Surface Defect Detection ◽  
The Mean

The existing seal ring surface defect detection methods for aerospace applications have the problems of low detection efficiency, strong specificity, large fine-grained classification errors, and unstable detection results. Considering these problems, a fine-grained seal ring surface defect detection algorithm for aerospace applications is proposed. Based on analysis of the stacking process of standard convolution, heat maps of original pixels in the receptive field participating in the convolution operation are quantified and generated. According to the generated heat map, the feature extraction optimization method of convolution combinations with different dilation rates is proposed, and an efficient convolution feature extraction network containing three kinds of dilated convolutions is designed. Combined with the O-ring surface defect features, a multiscale defect detection network is designed. Before the head of multiscale classification and position regression, feature fusion tree modules are added to ensure the reuse and compression of the responsive features of different receptive fields on the same scale feature maps. Experimental results show that on the O-rings-3000 testing dataset, the mean condition accuracy of the proposed algorithm reaches 95.10% for 5 types of surface defects of aerospace O-rings. Compared with RefineDet, the mean condition accuracy of the proposed algorithm is only reduced by 1.79%, while the parameters and FLOPs are reduced by 35.29% and 64.90%, respectively. Moreover, the proposed algorithm has good adaptability to image blur and light changes caused by the cutting of imaging hardware, thus saving the cost.

scholarly journals Research on a Surface Defect Detection Algorithm Based on MobileNet-SSD

2018 ◽  
Vol 8 (9) ◽  
pp. 1678 ◽  
Author(s):  
Yiting Li ◽  
Haisong Huang ◽  
Qingsheng Xie ◽  
Liguo Yao ◽  
Qipeng Chen
Keyword(s):  
Defect Detection ◽  
Surface Defect ◽  
Surface Defects ◽  
Detection Method ◽  
Detection Algorithm ◽  
Single Shot ◽  
Detection Model ◽  
Machine Learning Methods ◽  
Network Methods ◽  
Surface Defect Detection

This paper aims to achieve real-time and accurate detection of surface defects by using a deep learning method. For this purpose, the Single Shot MultiBox Detector (SSD) network was adopted as the meta structure and combined with the base convolution neural network (CNN) MobileNet into the MobileNet-SSD. Then, a detection method for surface defects was proposed based on the MobileNet-SSD. Specifically, the structure of the SSD was optimized without sacrificing its accuracy, and the network structure and parameters were adjusted to streamline the detection model. The proposed method was applied to the detection of typical defects like breaches, dents, burrs and abrasions on the sealing surface of a container in the filling line. The results show that our method can automatically detect surface defects more accurately and rapidly than lightweight network methods and traditional machine learning methods. The research results shed new light on defect detection in actual industrial scenarios.

scholarly journals Detection of Various Types of Metal Surface Defects Based on Image Processing

2021 ◽  
Vol 38 (4) ◽  
pp. 1071-1078
Author(s):  
Peng Xue ◽  
Changhong Jiang ◽  
Huanli Pang
Keyword(s):  
Image Processing ◽  
Metal Surface ◽  
Production Efficiency ◽  
Surface Defects ◽  
Detection Algorithm ◽  
Production Equipment ◽  
Single Shot ◽  
Target Image ◽  
Nonlocal Means ◽  
Surface Defect Detection

Machine vision is a promising technique to promote intelligent production. It strikes a balance between product quality and production efficiency. However, the existing metal surface defect detection algorithms are too general, and deviate from electrical production equipment in the level of response time to the target image. To address the two problems, this paper designs a detection algorithm for various types of metal surface defects based on image processing. Firstly, each metal surface image was preprocessed through average graying and nonlocal means filtering. Next, the principle of the composite model scale expansion was explained, and an improved EfficientNet was constructed to classify metal surface defects, which couples spatial attention mechanism. Finally, the backbone network of the single shot multi-box detector (SSD) network was improved, and used to fuse the features of the target image. The proposed model was proved effective through experiments.

scholarly journals Surface Defect Detection using Novel Histogram Distance-based Multiple Template Anomalies Detection Algorithm

2019 ◽  
Vol 7 (4.14) ◽  
pp. 401
Author(s):  
Ze-Hao Wong ◽  
C. M. Thong ◽  
W. M. Edmund Loh ◽  
C. J. Wong
Keyword(s):  
Deep Learning ◽  
Defect Detection ◽  
Surface Defect ◽  
Surface Defects ◽  
Detection Algorithm ◽  
Linear Filter ◽  
Wide Range ◽  
Surface Defect Detection ◽  
Design Variation ◽  
Industrial Lighting

Surface defects in manufacturing are top challenges in various manufacturing field including LED manufacturing, die manufacturing and printing industry. Quality control through automated surface defect detection has been an emphasis to speed up the production without jeopardizing the quality of the product. However, complexity and flexibility in product design, specification and dataset availability posted challenges in existing referential-based algorithm. Golden template-based algorithms are sensitive to misalignment and product variations. Deep learning and its variant can be used as non-linear filter to segment anomalies area. However, deep learning requires huge labelled database and consume long learning time. Similarly, maximum likelihood-based algorithms require large database for learning. This research proposes a novel histogram distance based multiple templates anomalies detection (MTAD) algorithm to segment surface defect. Histogram distance based on kernel-wise histograms stacked across illumination normalized database of similar size can describe the degree of anomaly intuitively across the image. Then, surface defect can be justified intuitively according to anomaly heat map generated. The algorithm is tested against industrial samples and it can handle texture and design variation existed in the product while catching anomaly in real time. This research suggests future studies on extending dimensionality of the histogram. Suggested algorithm has wide range of application other than surface defect detection. For examples, video motion detection, decolorization detection on industrial lighting.  

scholarly journals PSIC-Net: Pixel-Wise Segmentation and Image-Wise Classification Network for Surface Defects

Machines ◽  
2021 ◽  
Vol 9 (10) ◽  
pp. 221
Author(s):  
Linjian Lei ◽  
Shengli Sun ◽  
Yue Zhang ◽  
Huikai Liu ◽  
Wenjun Xu
Keyword(s):  
Deep Learning ◽  
Defect Detection ◽  
Surface Defect ◽  
Performance Metrics ◽  
Surface Defects ◽  
Training Data ◽  
Detection Methods ◽  
Detection Technology ◽  
Surface Defect Detection ◽  
Public Datasets

Recent years have witnessed the widespread research of the surface defect detection technology based on machine vision, which has spawned various effective detection methods. In particular, the rise of deep learning has allowed the surface defect detection technology to develop further. However, these methods based on deep learning still have some drawbacks. For example, the size of the sample data is not large enough to support deep learning; the location and recognition of surface defects are not accurate enough; the real-time performance of segmentation and classification is not satisfactory. In the context, this paper proposes an end-to-end convolutional neural network model: the pixel-wise segmentation and image-wise classification network (PSIC-Net). With the innovative design of a three-stage network structure, improved loss function and a two-step training mode, PSIC-Net can accurately and quickly segment and classify surface defects with a small dataset of training data. This model was evaluated with three public datasets, and compared with the most advanced defect detection methods. All the performance metrics prove the effectiveness and advancement of PSIC-Net.

A Collision Detection Algorithm Based on Spatial Partitioning and Bounding Volume

2013 ◽  
Vol 433-435 ◽  
pp. 932-935
Author(s):  
Wei Zhao ◽  
Li Ming Ye
Keyword(s):  
Real Time ◽  
Collision Detection ◽  
Detection Algorithm ◽  
Spatial Partitioning ◽  
The Real ◽  
Bounding Volume ◽  
Detection Technology ◽  
Bounding Boxes ◽  
Intersection Test ◽  
Rule Out

In order to improve the real-time and accuracy in the collision detection technology, a collision detection algorithm based on spatial partitioning and bounding volume was proposed . This algorithm adopted different spatial division strategies for different locations of the spaces according to the details in the scenes to exclude objects which can not intersect.Thus defined the potential intersection areas. Then we used a dynamic S-AABB hierarchy bounding boxes to test whether the intersection happened between the objects in the same grids. We used the sphere boxes to rule out the disjoint objects quickly. Then constructed the dynamic AABB bounding boxes trees for the rest of objects for further intersection test. At last, we improved the traditional overlapping test between the primitives for accurate collision detection . Compared to the traditional collision detection algorithm based on spatial partitioning and AABB bounding volume. This algorithm effectively improves the real-time of the collision detection without affecting the accuracy of original collision detection.

scholarly journals Virtual and Real-Time Synchronous Interaction for Playing Table Tennis with Holograms in Mixed Reality

Sensors ◽  
2020 ◽  
Vol 20 (17) ◽  
pp. 4857
Author(s):  
Bin Wang ◽  
Ruiqi Zhang ◽  
Chong Xi ◽  
Jing Sun ◽  
Xiaochun Yang
Keyword(s):  
Real Time ◽  
Collision Detection ◽  
Data Exchange ◽  
Mixed Reality ◽  
Detection Algorithm ◽  
Input Device ◽  
Table Tennis ◽  
Motion State ◽  
Detection Technology ◽  
Real Objects

Real-time and accurate interaction technology is required to realize new wearable Mixed Reality (MR) solutions. At present, the mainstream interaction method relies on gesture detection technology, which has two shortcomings: 1. the hand feature points may easily be obstructed by obstacles and cannot be detected and 2. the kinds of gesture that can be recognized are limited. Hence, it cannot support complex interactions well. Moreover, the traditional collision detection algorithm has difficulty detecting the collision between real and virtual objects under motion. Because location information of real objects needs updating in real time, it is easy to lose collision detection under high speeds. In the implementation of our system, Mixed Reality Table Tennis System, we propose novel methods which overcome these shortcomings. Instead of using gesture detection technology, we use a locator as the main input device and build a data exchange channel for the devices, so that the system can update the motion state of the racket in real time. Besides, we adjust the thickness of the collider dynamically to solve the collision detection problem and calculate rebound results responding to the motion state of the racket and the ball. Experimental results show that our method avoids losing collision detection and improves the authenticity of simulation. It keeps good interaction in real time.

Design of Bamboo Strip Detection System Based on Labview and Matlab Mixed Programming

2013 ◽  
Vol 302 ◽  
pp. 772-775 ◽  
Author(s):  
Li Ping He ◽  
Shu Xiang Song ◽  
Lei Liu ◽  
Xian Ming Jiang
Keyword(s):  
Real Time ◽  
Detection Rate ◽  
Surface Defects ◽  
Detection System ◽  
Detection Accuracy ◽  
The Novel ◽  
Large Area ◽  
Accuracy Requirement ◽  
Detection Technology ◽  
Mixed Programming

According to the domestic situation that bamboo strip detection technology is artificial intervention and low automation, the novel automatic bamboo strip detection system is designed based on Labview and Matlab mixed programming. The real-time detection on bamboo surface defects is realized in the system such as serious damage, stripe, scratch, large area dim. Moreover, the system provides more than 92% detection accuracy and eight strips per second detection rate, both of which can meet the accuracy requirement of practical production. In general, this paper provides a new kind of method for detection system.

scholarly journals Lightweight Fruit-Detection Algorithm for Edge Computing Applications

2021 ◽  
Vol 12 ◽  
Author(s):  
Wenli Zhang ◽  
Yuxin Liu ◽  
Kaizhen Chen ◽  
Huibin Li ◽  
Yulin Duan ◽  
...  
Keyword(s):  
Deep Learning ◽  
Real Time ◽  
Sampling Method ◽  
Feature Fusion ◽  
Detection Algorithm ◽  
Detection Accuracy ◽  
Model Structure ◽  
Excellent Performance ◽  
Backbone Network ◽  
Detection Technology

In recent years, deep-learning-based fruit-detection technology has exhibited excellent performance in modern horticulture research. However, deploying deep learning algorithms in real-time field applications is still challenging, owing to the relatively low image processing capability of edge devices. Such limitations are becoming a new bottleneck and hindering the utilization of AI algorithms in modern horticulture. In this paper, we propose a lightweight fruit-detection algorithm, specifically designed for edge devices. The algorithm is based on Light-CSPNet as the backbone network, an improved feature-extraction module, a down-sampling method, and a feature-fusion module, and it ensures real-time detection on edge devices while maintaining the fruit-detection accuracy. The proposed algorithm was tested on three edge devices: NVIDIA Jetson Xavier NX, NVIDIA Jetson TX2, and NVIDIA Jetson NANO. The experimental results show that the average detection precision of the proposed algorithm for orange, tomato, and apple datasets are 0.93, 0.847, and 0.850, respectively. Deploying the algorithm, the detection speed of NVIDIA Jetson Xavier NX reaches 21.3, 24.8, and 22.2 FPS, while that of NVIDIA Jetson TX2 reaches 13.9, 14.1, and 14.5 FPS and that of NVIDIA Jetson NANO reaches 6.3, 5.0, and 8.5 FPS for the three datasets. Additionally, the proposed algorithm provides a component add/remove function to flexibly adjust the model structure, considering the trade-off between the detection accuracy and speed in practical usage.

scholarly journals An Improved MB-LBP Defect Recognition Approach for the Surface of Steel Plates

2019 ◽  
Vol 9 (20) ◽  
pp. 4222 ◽  
Author(s):  
Yang Liu ◽  
Ke Xu ◽  
Jinwu Xu
Keyword(s):  
Real Time ◽  
Surface Defects ◽  
Recognition Accuracy ◽  
Detection System ◽  
Detection Algorithm ◽  
Steel Plates ◽  
Detection Accuracy ◽  
Scale Invariant ◽  
Speeded Up Robust Features ◽  
Real Time Detection

The detection of surface defects is very important for the quality improvement of steel plates. In actual production, as the steel plate production line runs faster, the steel surface defect detection algorithm is required to meet the requirements of real-time detection (less than 100 ms/image), and the detection accuracy is improved (at least 90%). In this paper, an improved multi-block local binary pattern (LBP) algorithm is proposed. This algorithm not only has the simplicity and efficiency of the LBP algorithm, but also finds a suitable scale to describe the defect features by changing the block sizes, thus ensuring high recognition accuracy. The experiment proves that the method satisfies the requirements of online real-time detection in terms of speed (63 ms/image), and surpasses the widely-used scale invariant feature transform (SIFT), speeded up robust features (SURF), gray-level co-occurrence matrix (GLCM), and LBP algorithms in recognition accuracy (94.30%), which prove that the MB-LBP has practical application value in an online real-time detection system.

Sign in / Sign up

Export Citation Format

Share Document