Showing papers on "Visual inspection published in 2016"

Machine learning-based imaging system for surface defect inspection

Abstract: Modern inspection systems based on smart sensor technology like image processing and machine vision have been widely spread into several fields of industry such as process control, manufacturing, and robotics applications in factories. Machine learning for smart sensors is a key element for the visual inspection of parts on a product line that has been manually inspected by people. This paper proposes a method for automatic visual inspection of dirties, scratches, burrs, and wears on surface parts. Imaging analysis with CNN (Convolution Neural Network) of training samples is applied to confirm the defect’s existence in the target region of an image. In this paper, we have built and tested several types of deep networks of different depths and layer nodes to select adequate structure for surface defect inspection. A single CNN based network is enough to test several types of defects on textured and non-textured surfaces while conventional machine learning methods are separately applied according to type of each surface. Experiments for surface defects in real images prove the possibility for use of imaging sensors for detection of different types of defects. In terms of energy saving, the experiment result shows that proposed method has several advantages in time and cost saving and shows higher performance than traditional manpower inspection system.

Machine vision system for food grain quality evaluation: A review

Abstract: Background Quality of pre-processed food grains is a critical aspect and a major decider of market acceptability, storage stability, processing quality, and overall consumer acceptance. Among various indices of food grain quality evaluation, physical appearance (including external morphology) provides the foremost assessment on the condition of the grain. Conventional method of grain quality evaluation, visual inspection (a manual method) is challenging even for trained personnel in terms of rapidity, reliability and accuracy. Scope and approach Machine vision systems have the potential to replace manual (visual) methods of inspection and, have therefore gained wide acceptance in industries as a tool for quality evaluation of numerous agricultural products. This note provides an up-to-date review on the major applications of machine vision systems for grain quality evaluation applications in non-touching arrangement, highlighting system components, image processing and image analysis techniques, advantages and limitations of machine vision systems. Key findings and conclusions Machine vision systems can provide rapid and accurate information about external quality aspects of food grains. However, it is a task to integrate such systems with those that can explain internal grain quality attributes. In the near future, with ever-growing application requirements and research developments, machine vision systems can provide effective solutions for various grain quality evaluation applications.

Comparison of Visual Inspection and Structural-Health Monitoring As Bridge Condition Assessment Methods

Abstract: This paper presents the results of a research project aimed at examining the capabilities and challenges of two distinct but not mutually exclusive approaches to in-service bridge assessment: visual inspection and installed monitoring systems. In this study, the intended functionality of both approaches was evaluated on its ability to identify potential structural damage and to provide decision-making support. Inspection and monitoring are compared in terms of their functional performance, cost, and barriers (real and perceived) to implementation. Both methods have strengths and weaknesses across the metrics analyzed, and it is likely that a hybrid evaluation technique that adopts both approaches will optimize efficiency of condition assessment and ultimately lead to better decision making.

Automated thermal fuse inspection using machine vision and artificial neural networks

Abstract: Machine vision is an excellent tool for inspecting a variety of items such as textiles, fruit, printed circuit boards, electrical components, labels, integrated circuits, machine tools, etc. This paper presents an intelligent system that incorporates machine vision with artificial intelligent networks to automatically inspect thermal fuses. An effective inspection flow is proposed to detect four commonly seen defects, including black-dot, small-head, bur, and flake during the production of thermal fuses. Backpropagation neural networks and learning vector quantization performance is compared in detecting the bur defect because of its illegibility. Different numbers of defective samples were screened out from a production line in a case study company and used to demonstrate the efficacy of the proposed system. Currently, the proposed inspection system is operating at the case study company, replacing four to six human inspectors. The system not only ensures the quality of the thermal fuses produced, but also reduced the cost of manual visual inspection.

Distortion-Free Image Mosaicing for Tunnel Inspection Based on Robust Cylindrical Surface Estimation through Structure from Motion

Abstract: Visual inspection, although labor-intensive, costly, and inaccurate, is a common practice used in the condition assessment of underground tunnels to ensure safety and serviceability. This p...

Vision-Based Corrosion Detection Assisted by a Micro-Aerial Vehicle in a Vessel Inspection Application

Abstract: Vessel maintenance requires periodic visual inspection of the hull in order to detect typical defective situations of steel structures such as, among others, coating breakdown and corrosion. These inspections are typically performed by well-trained surveyors at great cost because of the need for providing access means (e.g., scaffolding and/or cherry pickers) that allow the inspector to be at arm’s reach from the structure under inspection. This paper describes a defect detection approach comprising a micro-aerial vehicle which is used to collect images from the surfaces under inspection, particularly focusing on remote areas where the surveyor has no visual access, and a coating breakdown/corrosion detector based on a three-layer feed-forward artificial neural network. As it is discussed in the paper, the success of the inspection process depends not only on the defect detection software but also on a number of assistance functions provided by the control architecture of the aerial platform, whose aim is to improve picture quality. Both aspects of the work are described along the different sections of the paper, as well as the classification performance attained.

MHealth to Train Community Health Nurses in Visual Inspection with Acetic Acid for Cervical Cancer Screening in Ghana

Abstract: ObjectiveThere is a shortage of trained health care personnel for cervical cancer screening in low-/middle-income countries. We evaluated the feasibility and limited efficacy of a smartphone-based training of community health nurses in visual inspection of the cervix under acetic acid (VIA).Material

Quality Evaluation of Bakery Products

Abstract: Bakery products have many possible colors, shapes, and sizes. These attributes are influenced by many factors including the ingredients used and the processing environment. Damage during handling and packaging adds more kinds of colors, shapes, sizes, and other boundary irregularities. Therefore grading is an integral part in the bakery-producing industry. Traditionally, the quality assurance methods used in the food industry only involved human visual inspection. Hence, the decision is highly variable and the process is tedious, labor intensive, and subjective. The advent of computer-based technologies, such as machine vision, backed by powerful artificial intelligence has led to more automated inspection and improved process control. This chapter discusses powerful artificial classifiers and robust feature reduction techniques for the inspection of bakery products.

Vision-based control of a quadrotor for an object inspection scenario

Abstract: Unmanned aerial vehicles (UAVs) have gained special attention in recent years, among others in monitoring and inspection applications. In this paper, a less explored application field is proposed, railway inspection, where UAVs can be used to perform visual inspection tasks such as semaphore, catenary, or track inspection. We focus on lightweight UAVs, which can detect many events in railways (for example missing indicators or cabling, or obstacles on the tracks). An outdoor scenario is developed where a quadrotor visually detects a railway semaphore and flies around it autonomously, recording a video of it for offline post-processing. For these tasks, we exploit object detection methods from literature, and develop a visual servoing technique. Additionally, we perform a thorough comparison of several object detection approaches before selecting a preferred method. Then, we show the performance of the presented filtering solutions when they are used in servoing, and conclude our experiments with evaluating real outdoor flight trajectories using an AR.Drone 2.0 quadrotor.

Automated status inspection of fastening bolts on freight trains using a machine vision approach

Abstract: The status inspection and maintenance actions on the mechanical components of freight trains are significant determinants of railway safety. Fastening bolts, as a key component, are widely used to ...

An AR Inspection Framework: Feasibility Study with Multiple AR Devices

Abstract: We present an Augmented Reality (AR) based re-configurable framework for inspection that can be utilized in cross-domain applications such as maintenance and repair assistance in industrial inspection, health sector to record vitals, and automotive/avionics domain inspection, amongst others. The novelty of the inspection framework as compared to the existing counterparts are three fold. Firstly, the inspection check-list can be prioritized by detecting the parts viewed in inspector's field using deep learning principles. Second, the backend of the framework is easily configurable for different applications where instructions and assistance manuals can be directly imported and visually integrated with inspection type. Third, we conduct a feasibility study on inspection modes such as Google Glass, Google Cardboard, Paper based and Tablet for inspection turnaround time, ease, and usefulness by taking a 3D printer inspection use-case.

Vision-based fault inspection of small mechanical components for train safety

Abstract: With the development of both hardware and software technologies in camera and computer, automated visual inspection system is being used more and more in intelligent transportation system for its high efficiency. For the safety operation, it is necessary to perform fault inspection for train mechanical components. As one of the most widely used small mechanical components in freight trains, bogie block key (BBK) is used to keep wheel sets from separating out of bogies, and its fault is likely to cause terrible accidents. This study proposes a vision-based system to inspect the missing of BBK automatically. To ensure accurate and rapid fault inspection, a hierarchical detection framework consisting of fault area extraction and object detection is proposed. The purpose of fault area extraction is to divide image regions which contain the inspected component from the complex background. Subsequently, a component detector based on the sparse histograms of oriented gradients and support vector machine is proposed to verify the candidate image regions to check whether the BBK is missing or not. The experiments show that the proposed system realises the status inspection of BBK with high accuracy and high speed and can meet the need of actual applications.

MRIQC: Predicting quality in manual MRI assessment protocols using no-reference image quality measures.

Abstract: Quality control of MR images is essential for excluding problematic acquisitions and avoiding bias in subsequent image processing and analysis. However, the visual inspection of individual images is time-consuming and limited by both intra- and inter-rater variance. The difficulty of visual inspection scales with study size and with the heterogeneity of multi-site data. Here, we describe a tool for the automated assessment of T1-weighted MR images of the brain - MRIQC. MRIQC calculates a set of quality measures from each image and uses them as features in a binary (include/exclude) classifier. The classifier was designed to ensure generalization to new samples acquired in different centers and using different scanning parameters from our training dataset. To achieve that goal, the classifier was trained on the Autism Brain Imaging Data Exchange (ABIDE) dataset (N=1102), acquired at 17 locations with heterogeneous scanning parameters. We selected random forests from a set of models and pre-processing options using nested cross-validation on the ABIDE dataset. We report a performance of ∼89% accuracy of the best model evaluated with nested cross-validation. The best performing classifier was then evaluated on a held-out (unseen) dataset, unrelated to ABIDE and labeled by a different expert, yielding ∼73% accuracy. The MRIQC software package and the trained classifier are released as an open source project, so that individual researchers and large consortia can readily curate their data regardless the size of their databases. Robust QC is crucial to identify early structured imaging artifacts in ongoing acquisition efforts, and helps detect individual substandard images that may bias downstream analyses.

InspectAR: An Augmented Reality Inspection Framework for Industry

Abstract: With the advancement in camera technologies and data streaming protocols, AR based applications are proving to be an important aid for inspection, training and supervision tasks in various operations including automotive industry, education etc. We demonstrate an AR based re-configurable inspection framework that can be utilized in cross-domain applications such as maintenance and repair assistance in industrial inspection and automotive/avionics domain inspection, amongst others. A deep learning component detects parts viewed in inspector's Field-of-View (FoV) accurately and the corresponding inspection check-list can be prioritized based on detection results. The back-end of the framework is easily configurable for different applications where instructions can be directly imported and visually integrated with inspection type. Accurate recording of status of inspection is provided through evidence capturing of images, notes and videos. Our current framework supports all the Android based devices and will be demonstrated on Google Glass, Google Cardboard with smartphone, and Tablet with the help of 3D printer inspection use-case.

Vision-based robotic system for picking and inspection of small automotive components

Abstract: The use of vision systems for industrial robot guidance and quality control becomes much harder when the manufactured products and their components are small and possess reflective surface. To assure an effective automated visual inspection of such components, novel solutions are required, able to perform more advanced image analysis and tackle noise and uncertainty. This paper proposes a concept of multi-camera/multi-pose inspection station for star washers inspection, and presents the first results of a functional prototype implementation of it in a robotic cell. The processes of vision-guided part picking from a flexible feeder and close-range inspection in a dedicated rig are described. Solutions for the vision-based tasks of parts identification, machine learning-based classification, circular objects image analysis and star washer teeth segmentation are presented, and further directions are outlined.

Assessment of a Sampling Plan Based on Visual Inspection for the Detection of Anisakid Larvae in Fresh Anchovies ( Engraulis encrasicolus ). A First Step Towards Official Validation

Abstract: The presence of anisakid larvae in fish is a public health issue, and effective risk management procedures are needed to avoid that heavily infected products reach the market. Currently, an official sampling plan for fresh fish defining sample size, inspection methods, and criteria to accept or reject the merchandise is lacking at the European and Italian level. In this study, we compared the visual inspection proposed by the sampling plan of the Lombardy Region (Italy) to the UV press method and to an optimized digestion procedure with the aim to assess its ability in detecting visible parasites. Thirty-one batches of Engraulis encrasicolus, each composed of ∼30 specimens, were collected and subsequently analyzed with the three techniques. The mean abundance (MA) was calculated after each procedure and compared on the basis of a threshold value. The results showed that the visual inspection performed similarly to the digestion method, with a sensitivity of 93 %, a specificity of 100 %, and an accuracy of 97 %. Overall, the comparison showed that, in the proposed sampling plan, the visual inspection is effective in rejecting unmarketable anchovies and in preventing the commercialization of unsafe products. This method is simple, less demanding than digestion in terms of time and equipment, and thus suitable as a standardized procedure to be routinely applied by food business operators. The hazard characterization, performed by sequencing the mtDNA cox2 gene, has identified the visible larvae as Anisakis pegreffii in 98 % of the cases, highlighting the zoonotic potential of the parasites found and the need for preventive measures.

Visual Inspection of Vessels by Means of a Micro-Aerial Vehicle: An Artificial Neural Network Approach for Corrosion Detection

Abstract: Periodic visual inspection of the different surfaces of a vessel hull is typically performed by trained surveyors at great cost, both in time and in economical terms. Assisting them during the inspection process by means of mechanisms capable of automatic or semi-automatic defect detection would certainly decrease the inspection cost. This paper describes a defect detection approach comprising: (1) a Micro-Aerial Vehicle (MAV) which is used to collect images from the surfaces under inspection, particularly focusing on remote areas where the surveyor has no visual access; and (2) a coating breakdown/corrosion detector based on a 3-layer feed-forward artificial neural network. The success of the classification process depends not only on the defect detector but also on a number of assistance functions that are provided by the control architecture of the aerial platform, whose aim is to improve picture quality. Both aspects are described along the different sections of the paper, as well as the classification performance attained.

Scope and Prospects of Non-Invasive Visual Inspection Systems for Industrial Applications

Abstract: This research has been done to facilitate the industries and researchers with a new concept and emerging trends in the field of non-destructive visual inspections. Visual inspection is a process of determining the degree of deviation from a given set of specifications. And, visual inspection system (VIS) is a method of data acquisition, data analysis, quality control, electrical system control, and process control for a particular product, system or process. The main processes involved in a visual inspection system are: image acquisition, image de-noising, image enhancement, image segmentation, image feature selection and extraction, image classification, feature matching, decision making, display of results, and generation of controlling signals according to set values and parameters. The major challenges were identified while the designing and implementation of VIS are: illumination intensity, illumination angle, camera angle, selection of resolution of camera, selection of type of camera, setting of image acquisition rate, development of segmentation algorithm, and selection of an appropriate image processing techniques for a particular system. According to survey it is found that off line VIS based analysis is mostly being applied to analyze the quality of product. Although, non-conventional methods have got higher popularity than conventional methods but non-conventional methods are generally developed for particular applications and it is not flexible enough to be used for other similar inspections, on the other hand, the conventional methods such as: statistical approach and neural network based approach are reliable, flexible, and generalizable. The main categories of VIS were identified as On Line, Off Line and Real Time Visual Inspection Systems. The main area of VIS applications are: space craft inspection, railway track inspection, road traffic density inspection, rain forecasting, real time monitoring and controlling of automotive vehicles, inspection of turbines, and patient action monitoring.

A multicentre survey of the condition of ultrasound probes.

Abstract: BackgroundThe implementation of quality assurance for ultrasound scanners in the United Kingdom is patchy, but government appointed bodies require quality assurance and there are regulatory requirements for maintenance and inspection of equipment. Previous studies have shown high fault rates in ultrasound probes; some of these studies used electronic probe testers, but there is good evidence that over 90% of faults may be detected using simple methods. We aimed to conduct a multicentre survey of the condition of probes, using visual inspection and assessing the in-air reverberation.MethodsVisitors to the stand run by Multi-Medix Ltd at the BMUS Annual Scientific Meeting in 2014 were invited to participate in the study. One or both of the authors visited participants, performing a visual inspection of probes for evidence of damage or wear and inspecting the in-air reverberation pattern for uniformity. Probes were classified using a risk-based traffic light system: green—no fault found; amber—fault found; f...

An autonomous unmanned aerial vehicle sensing system for structural health monitoring of bridges

Abstract: As civil infrastructure (i.e. bridges, railways, and tunnels) continues to age; the frequency and need to perform inspection more quickly on a broader scale increases. Traditional inspection and monitoring techniques (e.g., visual inspection, mechanical sounding, rebound hammer, cover meter, electrical potential measurements, ultrasound, and ground penetrating radar) may produce inconsistent results, require lane closure, are labor intensive and time-consuming. Therefore, new structural health monitoring systems must be developed that are automated, highly accurate, minimally invasive, and cost effective. Three-dimensional (3D) digital image correlation (DIC) systems have the merits of extracting full-field strain, deformation, and geometry profiles. These profiles can then be stitched together to generate a complete integrity map of the area of interest. Concurrently, unmanned aerial vehicles (UAVs) have emerged as valuable resources for positioning sensing equipment where it is either difficult to measure or poses a risk to human safety. UAVs have the capability to expedite the optical-based measurement process, offer increased accessibility, and reduce interference with local traffic. Within this work, an autonomous unmanned aerial vehicle in conjunction with 3D DIC was developed for monitoring bridges. The capabilities of the proposed system are demonstrated in both laboratory measurements and data collected from bridges currently in service. Potential measurement influences from platform instability, rotor vibration and positioning inaccuracy are also studied in a controlled environment. The results of these experiments show that the combination of autonomous flight with 3D DIC and other non-contact measurement systems provides a valuable and effective civil inspection platform.

Development of SBC based machine-vision system for PCB board assembly Automatic Optical Inspection

Abstract: Visual Inspection is an essential part in the quality control in electronic manufacturing industry, especially in PCB assembly processes. With the advance of surface mount technology as a means to increase automation level in electronic assembly line, with production volume achieving thousands of board per day, manual visual inspection becomes increasingly prohitive. For this purpose, Automatic Optical Inspection (AOI) based on machine vision can be used. This method, however requires considerable amount of investment that is not affordable by small companies. In this research work, a low-cost machine vision based on an SBC (Single Board Computer) is developed. Raspberry-Pi™ as a populer SBC offers capability of being interfaced with camera and of running Linux-based operating system such as Ubuntu™. These capabilities enable this board to be used to run sophisticated image-processing programs for machine vision.

Development of Motion-Blur-Compensated High-Speed Moving Visual Inspection Vehicle for Tunnels

Abstract: In Japan, many infrastructures are several decades old or more, and since those structures are gradually deteriorating, efficient and precise monitoring methods are strongly required for maintaining safety. In particular, tunnels on highways must be monitored regularly; however, frequent traffic restrictions should be avoided. Accordingly, visual inspection of tunnels from a moving vehicle is an efficient method for rapidly discovering faults. However, despite the need for high image quality, motion blur deteriorates the image quality considerably, especially under high-speed motion. In the work described in this paper, we developed a motion-blur-compensated visual inspection system that uses a motion blur compensation method based on the back-and-forth motion of a galvanometer mirror. In field trials using a system installed on an actual vehicle, we confirmed the effect of motion blur compensation when using scales attached to the ceiling of a tunnel. The vehicle on which the inspection system was installed exceeded the minimum speed for Japanese highways, and the system was capable of distinguishing black-and-white stripes with widths of 0.2 mm. Additionally, this method can be used with conventional systems. 

Automatic visual inspection of a missing split pin in the China railway high-speed

Abstract: The split pin (SP) on the caliper brake is a vital component of the brake system of a bogie traveling along the China railway high-speed (CRH), and the absence of the SP could cause serious train accidents. A new automatic visual inspection method is proposed for the quick and accurate detection of SP faults of the CRH. The proposed approach is based on the histogram of gradient (HOG) combined with the complete local binary pattern (CLBP). First, a fast pyramid template matching technique is presented for localizing the region of interest to reduce the searching scope. Under the multiresolution pyramid model for target localization, a coarse-to-fine strategy is employed to ensure that the recognizing speed of the SP for the entire image is increased significantly. Second, a hierarchical framework is adopted at the localizing and inspecting stages of the SP to automatically implement the inspection tasks. To increase the robustness to the outside complex illumination, the HOG feature for localizing the target and the CLBP feature for examining the state of the SP (i.e., missing or not-missing) are extracted in the Sobel gradient domain. The localization and recognition stages are both fulfilled through the use of their respective intersection kernel support vector machine classifiers and corresponding features. In conclusion, experimental results indicate that the inspection system achieves a high accuracy rate of more than 99.0% and a real-time speed, thus proving that the proposed method is effective for the fault inspection of the SP and can satisfy the requirements of CRH’s actual application.

Development of a stable localized visual inspection system for underwater structures

Abstract: To avert potential crisis from Japan’s aging infrastructure and declining birth rate, the Japanese Government is planning to introduce robotic technology for the inspection of social infrastructure...

Monitoring and condition assessment of tramway track using in-service vehicle

Abstract: A uniform track condition assessment model, which is based on both visual inspection and automatic under load track geometry measuring system, is needed to reduce maintenance cost and increase safety and ride comfort for passengers. A tramway track condition assessment model as well as a geographical information system are worked out by the Author (implementation in progress) for Budapest tram lines to detect and predict rail defects and plan the effective maintenance work. The developing method determines the track condition on the basis of visual inspection and in-service vehicle’s wheels-mounted accelerometers.

Use of an eye-tracker to assess workers in ceramic tile surface defect detection

Abstract: The inspection and classification of ceramic tiles in a production line can be done automatically by use of digital cameras and image processing algorithms. However, due to the investment and maintenance costs of these systems, there are still several firms that assign workers for ceramic tile surface defect detection. Also, human inspection system can effectively challenge this task to a good accuracy. This study aims to attract attention to the mental workload that results from high concentration during visual inspection. A mobile type eye-tracker is used to record the data for duration of fixation and number of fixations to determine fatigue that arises over a period of working time. Data are analyzed and comments are made for workers, type of the ceramic tile, and working period without rest. It is concluded that the eye tracking systems have a potential to identify human related problems during visual inspection.

Collaborative Video Search Combining Video Retrieval with Human-Based Visual Inspection

Abstract: We propose a novel video browsing approach that aims at optimally integrating traditional, machine-based retrieval methods with an interface design optimized for human browsing performance. Advanced video retrieval and filtering (e.g., via color and motion signatures, and visual concepts) on a desktop is combined with a storyboard-based interface design on a tablet optimized for quick, brute-force visual inspection. Both modules run independently but exchange information to significantly minimize the data for visual inspection and compensate mistakes made by the search algorithms.