Machine Vision Inspection Research Articles

Research on Image Detection Technology of Printing Based on Dynamic Partitioning Method

Nowadays, many domestic printing enterprises still rely on artificial visual means to weed out waste, labor-intensive, time-consuming, subjective factors serious influence, the test results not being uniform, especially with the improvement of printing speed, gradually unable to meet the needs of the production. In order to solve the problems in the conventional assay and promote the development of China's printing industry productivity, there is an urgent need to develop the quality of printing online full screen detection system, with the rapid development of modern printing technology, the objective is to test and evaluate. In this paper, machine vision inspection technology, digital image processing technology is used in order to achieve the online testing and feedback of print quality. Dynamic partitioning method is based on standard proofs and proofs of defective segmentation and matching. According to the type of image resolution and the desired accuracy of detecting the type of design image segmentation method, thereby performing the detection and analysis of the color information. Secondary identification defect classification technology is used to change the defect classification, and then display the corresponding defect types. Compared with the traditional printing quality, the high detection print quality inspection speed and precision requirements have been reached, thereby the overall quality of the printed product is improved.

Development of an automatic image enhancement method using singular value decomposition for visual inspection

It is usually difficult to obtain good images for machine vision inspection in a manufacturing process. In practice, to obtain product characteristics, image enhancement methods are usually selected by trial-and-error or by experience. Therefore, image enhancement methods play a key role in image pre-processing. In this paper, we propose singular value decomposition (SVD) to extract the feature of images to automatically build image enhancement procedures. First, we completed image clustering according to the feature by using SVD. Next, the structural similarity index was used to select the optimal image enhancement method. To verify the procedures, 45 images from literature and local companies were used in the experimental analysis. For contrast value, the statistical analysis showed that the automatic enhancement result has no significant difference with the literature. The average entropy of the image relative to previous research increased to 17.54 %. The study results implied that the system could effectively improve the image quality and not over enhancement to produce noise.

Image analysis algorithm and verification for on-line molecular sieve size and shape inspection

An online machine vision inspection method is proposed to implement feedback control of molecular sieve growth process in rotary drum granulation. An experimental platform, comprising of a high-resolution digital camera and an image analysis system, has been developed to confirm the validity of the method on particle size distribution (PSD) and sphericity measurements. Experiments were performed with non-uniform molecular sieve particles (1–3mm diameter) obtained from production line. The particle images are first obtained through digital camera and are then processed by image analysis system. After separating the overlap particles and removing non-target particles of the images, the molecular sieve size and shape are computed in less than 0.9s. The validity of the size measuring accuracy is confirmed through comparing with the results from micrometer. The experimental results show that the repetitive precision of the proposed inspection system is about±1%, the diameter measurement error between image method and micrometer is about±3%, single image inspection speed is around 0.9s/frame. The proposed method is reliable to provide feedback information for control system in rotary drum granulation.

An Effective Defect Inspection Method for LCD Using Active Contour Model

Visual defects in liquid crystal display images often appear as low contrast and blurry contour without distinct intensity difference from their surrounding region. Besides, the background is usually intensity inhomogeneity. All these properties make the machine vision inspection extremely hard. This paper presents an effective machine vision inspection method using a local active contour model to detect defects with different brightness levels as well as diverse sizes and shapes. A modified local binary fitting model which is robust to initial contour is developed to extract defect boundary. Meanwhile, a simple preprocessing scheme is given to compensate the drawback of the two-phase active contour model for detecting objects with wide brightness levels. Experimental results show that the presented method can detect various types of defects effectively and achieves high performance in terms of inspection accuracy (both precision and recall are higher than 0.99).

The Study of Production Process System of New Multi-Feature Continuous Reel Soft Pack Trademarks

Continuous reel low weight and multi-color soft pack trademarks are used in the high-speed large-volume packaging production process during some new production process. However, most of the existing equipments in the factory are discrete single production machine. Therefore the production process of multi-feature continuous reel soft pack trademarks is needed to be solved urgently. Multi-feature continuous reel soft pack trademark set printing color clarity, color graded accuracy, trademarks pattern and text position accuracy in one. Disposable multi-point high-precision positioning technology, flat local heat transfer technology and digital intermittent control technology are used to achieve continuous reel trademarks high-precision double location synchronization hot stamping. That is, multiple convex word and laser holographic anti-counterfeit pattern intermittent synchronization hot stamping. Machine vision inspection system scan each pixel of trademark with high-speed CCD array and collect it into a computer database for comparison with the standard sample. The defects are recorded by electronic coordinates, building electron stripping system and removing unqualified trademarks, a set of production process system of new continuous reel soft pack trademarks is achieved. This research provides a theoretical basis for technological transformation of enterprises and products the new packaging production, completes its corresponding transformation and crafts integration. It could help enterprise form a new production process system.

Machine vision for crack inspection of biscuits featuring pyramid detection scheme

One of the challenges associated with machine vision inspection of biscuits or baked products with non-uniform colour distributions and textured background is the detection of a small and minute crack. In this study, a pyramid automatic crack detection scheme was proposed. This requires an enhancement method to properly distinguish the crack and intact samples. Canny–Deriche filter was used to emphasis the crack and reduce the noise. In order to segment minute crack pattern with less noise, a unimodal thresholding technique was developed and tested. The detection was based on support vector machine (SVM) featuring Wilk’s λ selection criteria. The accuracy of the system was compared with standard discriminant analysis. It was discovered that the pyramid SVM after Wilk’s λ analysis was more precise in detection compared to other classifiers, resulting in the specificity and sensitivity of 98% and 96% respectively, and average correct classification of consistently more than 97%.

A Low-Cost Color Vision System for Automatic Estimation of Apple Fruit Orientation and Maximum Equatorial Diameter

<abstract><title><italic>Abstract.</italic></title> The overall objective of this research was to develop an in-field presorting and grading system to separate undersized and defective fruit from fresh market-grade apples. To achieve this goal, a cost-effective machine vision inspection prototype was built, which consisted of a low-cost color camera, LED lights, and a generic bi-cone conveyor. Algorithms were developed for accurate estimation of pixels per unit dimension from images acquired under the close-range imaging configuration and for real-time estimation of apple orientation, shape, and maximum equatorial diameter. The machine vision system was tested and evaluated for ‘Delicious’ (D), ‘Empire’ (EM), ‘Golden Delicious’ (GD), and ‘Jonagold’ (JG) apples at a speed of four fruit per second. The variable pixels per unit dimension method achieved superior results for area estimation, compared to the conventional image distortion correction and area estimation methods. The orientation estimation algorithm had 87.6% and 86.2% accuracies for D and GD apples, respectively, within ±20° of actual fruit orientation, but it performed less satisfactorily for round-shaped EM and JG apples. The machine vision system achieved good fruit maximum equatorial diameter estimations, with an overall root mean squared error of 1.79 mm for the four varieties of apple, and it had a two-size grading error of 4.3%, versus 15.1% by a mechanical sizing machine. The system provides a cost-effective means for sorting apples for size.

Design of Nut Machine Vision Inspection and Screening Systems Based on PC

This system use PC as the core of image analysis and processing, with the single chip processor as the control core execution, combining with machine vision image processing technology, using advanced image processing algorithms, to achieve separation of the nut, and through experiments to test the correctness of the algorithm. The system has the advantage of a fast processing speed and high reliability. It not only save the manpower cost, but also improve the efficiency of the nut sorting.

Construction of an automation system for the inspection and packing processes of a screw/bolt production line

In this paper, a packing automation system was constructed and a prototype was manufactured to test its performance for the inspection and packing process automation of a screw/bolt production line. The packing automation system is composed of an automated inspection system and an automated packing system. An inspection system using machine vision inspection and the slant method was proposed, developed and applied. The system was also developed and applied as the MICOM system in the PLC system to improve inspection errors due to the delays in the inspection process. The feeder to supply work to the inspection device was proposed, developed and applied as a slide feeder by applying a vertical transfer mechanism. The slide method was designed to minimize friction and thus improve upon the disadvantages of the existing bowl feeder. An automated packing system based on a one-station method for two-sided vinyl automatic packing was proposed, developed and applied. The one-station method maximizes the speed of the packing processes, such as automatic supply, cutting, spreading, charging, closing and sealing, that occur over a short distance. Performance of the system was evaluated by producing a prototype and setting up a packing automation system. The performance test was validated based on the inspection handling speed, transfer speed of the slide feeder, packing speed of the automatic packing system and charging accuracy, verifying the applicability and practicality of the automated system for the screw/bolt production line.

Review on Key Technology of the Machine Vision Inspection Robot on Pharmaceutical Injection

The food and drug safety is related to people’s livelihood and public safety issues,so medical injection must be through quality and safety tested when using.Otherwise,it will cause a variety of diseases or even life-threatening.At present,pharmaceutical manufacturers still use artificial lighting inspection method to control the quality of injection.This inspection method has many defects,such as low speed,poor stability,low level of security.So in order to solve this problem and improve the level of drug security status,it is very necessary to introduce a type of robot based on machine vision technology to inspect pharmaceutical injection instead of people.Key technical problems encountered are discussed in the development process of the inspection robot,including overall solutions of the robot,the design of high precision mechanical structure,the synergistic control method of sever motors,the high-speed and high-precision image acquisition system and methods of weak target to detect,recognize and classify.The research and development bottlenecks which constraint the robot and feasible solutions are elaborated,and the next step in the development trend of the visual inspection robot is stated.

Remote-Controlled Visual Inspection Machine for Bridge Supports

We have developed a remote-controlled visual inspection machine for bridge supports. This inspection machine adheres to concrete structures using two sets of suction cups, and ascends by alternating movements of the two sets of suction cups. Attached to the inspection machine is a CCD video camera. Video data of the condition of bridge supports is transmitted wirelessly to the operator. The visual inspection machine can ascend vertical concrete structures and then inspect bridge supports by moving horizontally. The maximum height the machine is capable of inspecting is limited by the length of the power cables as well as the weight. It will be able to inspect higher bridge supports when the weight of power cables can be reduced. The weight and the dimensions of the testing machine are about 17 kg and 84 cm by 72cm, respectively

Reconfigurable Image Acquisition in the Machine Vision Inspection System

Since camera drivers provided by the different manufacturers are not compatible, machine vision systems must be redeveloped according to specific camera. It is great significant to work out the problem, which could improve the versatility of the inspection system. The reconfigurable technology has applied to image processing, image matching and so on. Hence, in the paper the reconfigurable image acquisition module is designed, which reserves some interfaces for the image detection module. Citing the nonel visual inspection system as an example, adopting DALSA and BASLER cameras to acquire the images, the images was displayed properly. Therefore, the compatibility of the image detection system has been improved greatly.

Design of Automatic Vision-based Inspection System for Monitoring in an Olive Oil Bottling Line

vision inspection has become a vital part of the quality monitoring process. This paper compares the development and performance of two methodologies for a machine vision inspection system online for high speed conveyor. The first method developed is the Thresholding technique image processing algorithms and the second method is based on the edge detection. A case study was conducted to benchmark these two methods. Special effort has been put in the design of the defect detection algorithms to reach two main objectives: accurate feature extraction and on-line capabilities, both considering robustness and low processing time. An on-line implementation to inspect bottles is reported using new communication technique with GigE Vision camera and industrial Gigabit Ethernet network. The system is validated on olive oil bed. The implementation of our algorithm results in an effective real-time object tracking. The validity of the approach is illustrated by the presentation of experiment results obtained using the methods described in this paper.

Automatic Inspection System of Small Plastic Gears Based on Machine Vision

Small plastic gear is generally made by injection molding.But the injection molding process and mold have problems with missing tooth, shrink, more material, less material and inaccurate roundness and so on. Furthermore, using manual inspection will appear phenomenon of low efficiency, false detection and leak detection. To solve these problems, this paper introduces an automatic inspection system of small plastic gears based on machine vision. The system consists of feeding and sorting machine control system and machine vision inspection system of the gear defects. Mechanical control use digital servo control technology to achieve automatic nesting, feeding, positioning of gear workpiece, and depend on the inspection result of machine vision system to sort. After acquired gear image through a camera, Machine vision system uses median filtering, binarization, edge detection algorithms to process image. Then the system adopts template matching algorithm to obtain the inspection result and send the result to the sorting controller, which achieve automatic smart inspection of gear. The automatic inspection system has accurate, efficient, intelligent and other advantages.

Research of the Paper Defect On-Line Inspection System Based on Distributed Machine Vision

This paper presents a distributed machine vision inspection system, which has a large field of view (FOV) and can perform high precision, high speed real-time inspection for wide paper sheet detection. The system consists of multiple GigE Vision linescan cameras which connected though Gigabit Ethernet. The cameras are arranged into a linear array so that every camera’s FOV is merged into one large FOV in the meantime the resolution keeps unchanged. In order to acquire high processing speed, the captured images from each camera are sent into one dedicate computer for distributed and parallel image processing. Experimental results show that the system with fine detection capability can satisfy the requirements of real time detection and find out the defects on the production line effectively.

Research on Dimensional Inspection System of Thin Sheet Part Based on Machine Vision

Key technologies of dimensional inspection system of thin sheet part based on machine vision are investigated, and an entire machine vision inspection system is developed. A cad information-based line scanning step adaptive optimization method used for image grabbing of inspected part is proposed. A rectangle lens subpixel edge detection method based on cubic spline interpolation used for edge detection is advanced. A planar contour primitive recognition method based on curvature and HOUGH transform used for image recognition is raised. The inspection accuracy of the inspection system can reach to 1μm, and the inspection time can satisfy the requirements of on-line real-time inspection, so the inspection system is feasible.

Computational Model for Machine Vision Inspection Based on Vision Attention

Human vision system exploits this fact by visual selective attention mechanisms towards important and informative regions. A computational model of combination both bottom-up and top-down simulating human vision system for machine vision inspection is proposed. In this model, top-down knowledge-based information is highlighted to integrate into bottom-up stimulus-based process of vision attention. The model is tested on inspecting contaminants in cotton images. Experiment result shows that the proposed model is feasible and effective in visual inspection. And it is available and quasi-equivalent to human vision attention.

An Online Visual Inspection System for Oil-Seal Dimension

Adoption of machine vision inspection and computer image processing technology, an oil-seal dimension measuring system was developed to meet the requirement of online production and real time inspection. The makeup and principle of the system were introduced, as well as its working process and design requirements were described on detailed. The technique of quadratic filtering for image preprocessing combined with the principle of three points determining a circle, point Hough transform and the least squares was employed for image processing algorithm, and high precision sub-pixel edge detection was achieved. The measuring results of experiments demonstrated that the inspection goal on 100 percents of products could be realized successfully, and with many advantages such as non-contact, on-line, real time, appropriate precision and low cost, the system can be applied widely in other production fields.

Temporal Feature Characterization via Dynamic Hidden Markov Tree

We present a novel multiscale dynamic methodology for automatic machine vision inspection aiming at characterizing temporal features of tobacco leaves. The image sequences of tobacco leaves are transformed from RGB color space to L*a*b* color space, which provides a uniform perceptual difference measure. The image sequences are then represented by a multiscale Dynamic Hidden Markov tree (DHMT), which models not only inter and intra scale dependences of wavelet coefficients, but also temporal dependences of foreground/background observational properties. Experimental results demonstrate temporal consistent mean and covariance values of model coefficients in a* color channel.

Multivariate analysis-based image enhancement model for machine vision inspection

Image enhancement is an essential procedure in machine vision-based inspection. In practical applications, image enhancement is usually a part of image pre-processing, intended to make the following inspection more effective. The image enhancement method is usually selected by trial-and-error or on the basis of experience. This paper presents an automatic procedure for fast and effective image enhancement. The procedure uses multivariate analysis to automatically construct an optimal image enhancement model. First, an optimally enhanced image was selected from the literature as a basis for the model. Then, the image features were identified and Wilks’ statistic was used for feature selection. Next, discriminate functions were built to select the optimal image enhancement method. To verify the model, 53 training images from the literature and 12 test images from a local company were used in an experimental analysis. The model achieved 98.11% accuracy in selecting the most suitable image enhancement method, and the average increase in contrast was 98% for the 53 training images. The enhancement method selection results for the 12 test images were also in agreement with the 53 training images from the literature. The results show that the proposed method is effective and appropriate for quickly improving image contrast.