Defect Inspection Algorithm Research Articles

An automatic machine vision-based algorithm for inspection of hardwood flooring defects during manufacturing

Hardwood flooring products are popular construction materials because of their aesthetics, durability, low maintenance requirements, and affordability. To ensure product quality during manufacturing, common defects such as cracks, chips, or stains are typically detected and classified manually, but this process can decrease productivity. The aim of this study was to develop an automatic machine vision-based inspection system with a robust algorithm for inspecting small hardwood flooring defects in a production line. This defect-inspection algorithm is based on image-processing techniques, including background elimination, boundary approximation, and defect inspection of photographs. The YOLOv5 deep-learning algorithm for object detection was applied to detect surface defects. The resulting algorithm identified the quality of each specimen (i.e., either good or defective). The influences of colour and surface patterns on defect inspection were experimentally investigated under light conditions. The algorithm was adaptable to specimens with different colours and patterns under various conditions, demonstrating the potential of this approach in practical situations.

An Efficient YOLO Algorithm with an Attention Mechanism for Vision-Based Defect Inspection Deployed on FPGA.

Industry 4.0 features intelligent manufacturing. Among them, the vision-based defect inspection algorithm is remarkable for quality control in parts manufacturing. With the help of AI and machine learning, auto-adaptive instead of manual operation is achievable in this field, and much progress has been made in recent years. In this study, considering the demand of inspection features in industrialization, we made further improvement in smart defect inspection. An efficient algorithm using Field Programmable Gate Array (FPGA)-accelerated You Only Look Once (YOLO) v3 based on an attention mechanism is proposed. First, because of the relatively fixed camera angle and defect features, an attention mechanism based on the concept of directing the focus of defect inspection is proposed. The attention mechanism consists of three improvements: (a) image preprocessing, which is to tailor images for selectively concentrating on the defect relevant things. Image preprocessing mainly includes cutting, zooming and splicing, named CZS operations. (b) Tailoring the YOLOv3 backbone network, which is to ignore invalid inspection regions in deep neural networks and optimize the network structure. (c) Data augmentation. First, two improvements can be made to efficiently reduce deep learning operations and accelerate the inspection speed, but the preprocessed images are similar and the lack of diversity will reduce network accuracy. So, (c) is added to mitigate the lack of considerable amounts of training data. Second, the algorithm is deployed on a PYNQ-Z2 FPGA board to meet the industrialization production requirements for accuracy, efficiency and extensibility. FPGA can provide a low-latency, low-cost, high-power-efficiency and flexible architecture that enables deep learning acceleration for industrial scenarios. A Xilinx Deep Neural Network Development Kit (DNNDK) converted the improved YOLOv3 to Programmable Logic (PL), which can be deployed on FPGA. The conversion process mainly consists of pruning, quantization and compilation. Experimental results showed that the algorithm had high efficiency, inspection accuracy reached 99.2%, processing speed reached 1.54 Frames per Second (FPS), and power consumption was only 10 W.

Real-time defect inspection of green coffee beans using NIR snapshot hyperspectral imaging

Coffee beans are important agricultural commodities traded in the international market. Screening for defective beans is an important step before roasting. The main types of defective beans include black, fermented, moldy, insect damaged, shell, and broken. Insect-damaged beans are the most common type of defective beans. Previously, coffee beans were sorted manually, which was extremely labor intensive and prone to fatigue-induced errors, resulting in inconsistent quality. This study combines a near-infrared snapshot hyperspectral sensor and deep learning to create a multimodal real-time coffee-bean defect inspection algorithm (RT-CBDIA) for sorting defective green coffee beans. Furthermore, three convolutional neural networks (CNN) were designed to achieve real-time inspection, i.e., lean 2D-CNN, 3D-CNN, and 2D–3D-merged CNN. Subsequently, principal component analysis was used to select important bands. Our experimental results achieved an overall accuracy of 98.6% using 1026 green coffee-bean samples. Furthermore, the RT-CBDIA achieved a Kappa value of 97.2% and real-time sorting speeds. These achievements are considerably beneficial for subsequent applications and the commercialization of smart agriculture. Our main objective is to commercialize the proposed RT-CBDIA algorithm by combining it with a robot to create a comprehensive yet affordable coffee-bean real-time inspection system. It can be used to achieve real-time and noninvasive inspections while reducing labor costs. In the future, our real-time inspection system can also be applied to other crops to ultimately advance smart agriculture.

Deep-learning-based anomaly detection for lace defect inspection employing videos in production line

Defect inspection plays an essential role in ensuring quality of industrial products. The most widely used human visual inspection method has some drawbacks such as high cost and low efficiency, which bring an eager demand for the application of automatic defect inspection algorithm in actual production. However, few industrial production lines use automatic detection devices due to the gap between data collected in the actual production environment and ready-made datasets. Lace is one of the industrial products which completely depends on manual defect inspection. The complex and fine texture of lace makes it difficult to extract regular patterns using the existing image-based defect inspection methods. In this paper, we propose to collect lace videos in the weaving stage and design a deep-learning-based anomaly detection framework to detect lace defects. The framework contains three stages, namely video pre-processing stage, pixel reconstruction stage and pixel classification stage. In the offline phase, only defect-free lace videos are needed to train the pixel reconstruction model and calculate the detection threshold by our adaptive thresholding method. In the online phase, the proposed framework reconstructs lace videos and performs defect inspection using reconstruction error and the pre-set threshold. As far as we know, this paper the first to detect fabric defects by videos. Experimental results on artificial defect videos demonstrate the effectiveness of the proposed framework.

Unsupervised defect inspection algorithm based on cascaded GAN with edge repair feature fusion

Unsupervised defect inspection algorithm based on cascaded GAN with edge repair feature fusion

Preface

It is our pleasure to welcome you to 2021 2nd International Workshop on Electronic communication and Artificial Intelligence (IWECAI 2021) in Nanjing, China during March 12-14, 2021. IWECAI is the premier forum that brings both industry and academia for the dissemination of technological advances, research results and applications in the fields of Artificial Intelligence Methods and Technologies. The conference provided a valuable opportunity for researchers and industry experts to exchange their ideas face to face.The conference model was divided into two sessions, including oral presentations and keynote speeches. In the first part, some scholars, whose submissions were selected as the excellent papers, were given 15 minutes to perform their oral presentations one by one. Then in the second part, keynote speakers were each allocated 30-45 minutes to hold their speeches. There were over 110 experts and scholars in the area of Electronic communication and Artificial Intelligence representing different famous universities and institutes around the globe to attend the conference.During the conference, we were pleased to invite two distinguished experts to present their insightful speeches. Nikhil R. Pal, our first keynote speaker, from National Academy of Sciences of India. His research area included brain science, computational intelligence, machine learning and data mining. Dr. Tian Qiu, from Wuyi University, China. He talked about A liquid-lens based defect inspection algorithm for optical isolators. The experiments show that the algorithm achieves satisfactory results in the detection inspection, which can reduce the manpower demand of the visual inspection to less than 10%.The participants of the conference were from most part of the world, with background of either academia or industry, even well-known enterprise. The success and prosperity of the conference is reflected high level of the papers received. The proceedings are a compilation of the accepted papers and represent an interesting outcome of the conference. IWECAI 2021 for exchange of research findings, analysis, information, and knowledge in areas which include but are not limited to: Information and Communication Engineering, Electronics and Communications Engineering, Intelligent System, Artificial Intelligence & Applications and other related topics.We would like to acknowledge all of those who supported IWECAI 2021. Each individual and institutional help were very important for the success of this conference. Especially we would like to thank the organizing committee for their valuable advices in the organization and helpful peer review of the papers. We sincerely hope that IWECAI 2021 will be a forum for excellent discussions that will put forward new ideas and promote collaborative researches. We are sure that the proceedings will serve as an important research source of references and the knowledge, which will lead to not only scientific and engineering progress but also other new products and processes.The Organizing Committee of IWECAI 2021Committee member are available in the pdf

A steel surface defect inspection approach towards smart industrial monitoring

With the advance in Industry 4.0, smart industrial monitoring has been proposed to timely discover faults and defects in industrial processes. Steel is widely used in manufacturing equipment, and steel surface defect inspection is of great significance to the normal operation of steel equipment in manufacturing workshops. In steel defect inspection systems, industrial inspection robots generate images via scanning steel surface, and processors perform surface defect inspection algorithms on images. We focus on applying advanced object detection techniques to surface defect inspection algorithm for sheet steel. In the proposed steel surface defect inspection model, a deformable convolution enhanced backbone network firstly extracts complex features from multi-shape steel surface defects. Then the feature fusion network with balanced feature pyramid generates high-quality multi-resolution feature maps for the inspection of multi-size defects. Finally, detector network achieves the localization and classification of steel surface defects. The proposed model is evaluated on a typical steel surface defect dataset. Our model achieves 0.805 mAP, 0.144 higher than baseline models, and our model shows high efficiency in inference. Experiments are performed to reveal the effect of employed approaches, and results also show our model achieves a balance between inspection performance and inference efficiency.

Automated Defect Inspection Algorithm Incorporated Spectral-Domain Optical Coherence Tomography for Optical Polarizing Thin Films

Optical polarizing thin film is an optical filter enables light waves of a specific polarization pass through while blocking light waves of other polarizations. Optical polarizing thin films control the brightness of back-light unit for LCD (liquid crystal display) panel, which is essential to produce LCD modules. Defect inspection of polarizing thin films is an important feature during the manufacturing process that is helpful to improve the product quality. In the current study, an automated defect inspection algorithm is introduced and incorporated with a well-known non-destructive and non-contact optical inspection method called spectral domain optical coherence tomography (SD-OCT) to pre-identify defective sub-surface as well as top-surface locations of optical polarizing thin films Polarizing thin films employed in this study consist of 6 layers. The tomographic information, layer information, and defective locations were sufficiently identified through the SD-OCT system owing high-axial resolution. The acquired results indicate the possible application of the proposed system in optical polarizing thin films for the quality assurance.

Development of an optical defect inspection algorithm based on an active contour model for large steel roller surfaces

On-site measurements and defect detection are of great importance for precision ground steel rollers due to their large dimension and weight. In addition to dimensional error, form accuracy, surface roughness, and surface/sub-surface cracks, there also exist optical defect requirements for steel roller surfaces, e.g., speckles, chatter marks, or feed traces. Since rollers with optical defects will always duplicate the defect patterns onto the metal sheet or foil during rolling, it is necessary as well as significant to scrutinize the roller surface after grinding. In industrial practice, defects are investigated mainly by experienced engineers through naked-eye inspections along particular directions and under appropriate illumination conditions. This is usually subjective and inconsistent. In this paper, a machine vision system is developed, to add onto the roller grinder, that is capable of acquiring the roller's surface image with high and consistent quality. In addition, to identify defects with fuzzy boundaries, intensity inhomogeneity, and complex background textures, an improved segmentation algorithm is developed based on an active contour without edges model. Furthermore, qualitative and quantitative comparisons of the proposed algorithm with the Chan-Vese model, the local binary fitting model, and the globally signed region pressure force model are carried out. The comparisons prove that the proposed method performs with better accuracy and robustness for fuzzy and inhomogeneous defect segmentation and consumes generally less computational time.

A novel surface defect inspection algorithm for magnetic tile

In this paper, we propose a defect extraction method for magnetic tile images based on the shearlet transform. The shearlet transform is a method of multi-scale geometric analysis. Compared with similar methods, the shearlet transform offers higher directional sensitivity and this is useful to accurately extract geometric characteristics from data. In general, a magnetic tile image captured by CCD camera mainly consists of target area, background. Our strategy for extracting the surface defects of magnetic tile comprises two steps: image preprocessing and defect extraction. Both steps are critical. After preprocessing the image, we extract the target area. Due to the low contrast in the magnetic tile image, we apply the discrete shearlet transform to enhance the contrast between the defect area and the normal area. Next, we apply a threshold method to generate a binary image. To validate our algorithm, we compare our experimental results with Otsu method, the curvelet transform and the nonsubsampled contourlet transform. Results show that our algorithm outperforms the other methods considered and can very effectively extract defects.

Automatic optical inspection system for IC molding surface

Success or failure of an IC product hinges on the quality of molding process which protects chips from the harm done by external force and moisture. Defects such as cracks, dilapidations or voids may be embedding on the molding surface while a chip was being molded. Human inspection often neglects a very tiny crack or a low-contrast void. Hence an automatic optical inspection system for the integrated circuit (IC) molding surface cannot be over emphasized. The proposed system is composed of a charged coupled device, a coaxial light, a back light and a motion control unit. Based on the characteristics of statistical textures of the molding surface, a series of digital image processing is carried out, including normalization, shrinking, segmenting and Fourier based image restoration and defect identification. Training of the parameter associated with defect inspection algorithm is also discussed. Results of the experiment suggest that the inspection system works effectively with high accuracy rate of 94.2 %, contributing to the inspection quality of IC molding surface.

Surface Casting Defects Inspection Using Vision System and Neural Network Techniques

Abstract The paper presents a vision based approach and neural network techniques in surface defects inspection and categorization. Depending on part design and processing techniques, castings may develop surface discontinuities such as cracks and pores that greatly influence the material’s properties Since the human visual inspection for the surface is slow and expensive, a computer vision system is an alternative solution for the online inspection. The authors present the developed vision system uses an advanced image processing algorithm based on modified Laplacian of Gaussian edge detection method and advanced lighting system. The defect inspection algorithm consists of several parameters that allow the user to specify the sensitivity level at which he can accept the defects in the casting. In addition to the developed image processing algorithm and vision system apparatus, an advanced learning process has been developed, based on neural network techniques. Finally, as an example three groups of defects were investigated demonstrates automatic selection and categorization of the measured defects, such as blowholes, shrinkage porosity and shrinkage cavity.

A Template Model for Defect Simulation for Evaluating Nondestructive Testing in X-Radiography

This paper proposes a new template model for the simulation of casting defects which are classified, according to shape, into three main types: the single defect with circular or elliptical shape, shrinkage defects with stochastic discontinuities, and the cavity or sponge shrinkage defects. For effective simulation, different nesting stencil plates are designed to reflect the characteristics of different casting defects. These include intensity, orientation, size, and shape. The proposed approach also uses geometric diffusion to demonstrate the production of simulated defects with effective shading and contrast when compared to their background. In order to evaluate the effectiveness of the proposed approach, the simulated casting defects are superposed on real radioscopic images of casting pieces and compared with real defects by extensive visual inspection. On the other hand, in order to verify the similarity of the simulated defects and the real defects, we have used our defect inspection algorithm to recognize both real and simulated defects in the same image. The experimental results show that the proposed defect simulation approach can produce a large range of simulated casting defects, which can be utilized as sample images to tune the parameters of casting inspection algorithms.

Machine Vision System for Automatic Inspection of Surface Defects in Aluminum Die Casting

A camera based machine vision system for the automatic inspection of surface defects in aluminum die casting is presented. The system uses a hybrid image processing algorithm based on mathematic morphology to detect defects with different sizes and shapes. The defect inspection algorithm consists of two parts. One is a parameter learning algorithm, in which a genetic algorithm is used to extract optimal structuring element parameters, and segmentation and noise removal thresholds. The second part is a defect detection algorithm, in which the parameters obtained by a genetic algorithm are used for morphological operations. The machine vision system has been applied in an industrial setting to detect two types of casting defects: parts mix-up and any defects on the surface of castings. The system performs with a 99% or higher accuracy for both part mix-up and defect detection and is currently used in industry as part of normal production.