Machine vision system for curved surface inspection

Lee, Min‐Fan Ricky; Silva, Clarence W. de; Croft, Elizabeth A.; Wu, Qi

doi:10.1007/s001380000043

Cited by 21 publications

(9 citation statements)

References 2 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Due to the uncertainties in the classifier performance, Hu et al [11] also propose a general grading framework that includes manual regrading of the fraction of roe that lies near the classification border between 'good' and 'poor'. Later versions of the roe grading system included color features, and 3D imaging using multiple laser stripes was added in order to enable automated weighing of the roe and detection of deformed (3D deformations) class of roe called 'henkei' [14]. This previous work is topically close to ours, even though it does not distinguish between herring fractions.…”

Section: Introductionmentioning

confidence: 83%

A Machine Vision System for Robust Sorting of Herring Fractions

Guttormsen

Toldnes

Bondø

et al. 2016

Food Bioprocess Technol

View full text Add to dashboard Cite

ABSTRACT:19 Among the rest raw material in herring (Clupea harengus) fractions, produced during the 20 filleting process of herring, there are high value products such as roe and milt. As of today there 21 has been little or no major effort to process these by-products in an acceptable state, except 22 for by manual separation and mostly mixed into low-value products. Even though pure roe and 23 milt fractions can be sold for as much as ten times the value of the mixed fractions, the 24 separation costs using manual techniques render this economically unsustainable. Automating 25 this separation process could potentially give the pelagic fish industry better raw material 26 utilization and a substantial additional income. In this paper, a robust classification approach is 27 described which enables separation of these by-products based on their distinct reflectance

show abstract

Section: Introductionmentioning

confidence: 83%

A Machine Vision System for Robust Sorting of Herring Fractions

Guttormsen

Toldnes

Bondø

et al. 2016

Food Bioprocess Technol

View full text Add to dashboard Cite

show abstract

“…Optical methods, such as interferometry [20], stereo vision [21][22][23], structured light projection [24,25] and shape from focus/defocus [26][27][28] techniques, have long received extensive attention for reconstruction of 3D objects. Unlike CMM, measurement systems developed from optical principles are non-contact in nature, and consequently mechanical wear and surface deformation problems do not appear.…”

Section: Overviewmentioning

confidence: 99%

Accurate 3D Reconstruction using a Turntable-based and Telecentric Vision

González-Barbosa

Rico-Espino²,

Gómez-Loenzo

et al. 2015

Automatika

View full text Add to dashboard Cite

Original scientific paperThis paper presents a camera+telecentric lens that is able to obtain 3D information. We designed and implemented a method which can register and integrate 3D information captured from different viewpoints to build a complete 3D object model. First, a geometric model of a camera+telecentric lens is established. Then a calibration process using a planar checkerboard is developed and implemented. The object is placed on a rotation stage in front of a stationary camera. Normally the rotation axis is considered to be aligned with camera frame. In the description presented in this paper, the rotation matrix and translation vector of the rotation axis are calibrated. At the same time, a three-dimensional reconstruction system based on contour extraction of objects with dimensions less than 50 mm in diameter is developed. Finally, an analysis of the uncertainty model parameters and performance reconstruction of 3D objects are discussed.Key words: Calibration, 3D Reconstruction, Telecentric lens Točna 3D rekonstrukcija zasnovana na rotirajućoj platformi i telecentričnoj viziji.Članak prestavlja sustav koji se sastoji od kamere i telecentrične leće koji omogućavaju dobivanje 3D informacije o objektu. Dizajnirana je i implementirana metoda koja može registrirati i integrirati 3D informacije iz različitih točaka gledišta, kako bi se izgradio potpuni 3D model. Na početku, uspostavlja se geometrijski model kamere i telecentrične leće. Nakon toga koristi se razvijena metoda kalibracije zasnovana na šahovskoj ploči te se objekt postavlja na rotirajuću platformu ispred stacionarne kamere. Takoîer, pretpostavlja se da je os rotacije poravnta s koordinantim sustavom kamere. U ovomečlanku kalibriraju se rotacijska matrica i translacijski vektor rotacijske osi. Razvijen je i sustav 3D rekonstrukcija zasnovan na izlučivanju kontura objekta dimenzija manjih od 50 mm u promjeru. Na kraju, provedena je i analiza nesigurnosti parametara modela kao i točnost rekonstrukcije 3D modela.

show abstract

“…This increases the number of images captured and thus proportionately increases the total time for inspection. The major challenges are the highly reflective nature of chrome coated surfaces (Khalili and Webb, 2007;Rosati et al, 2009), the curved nature of the surface to be inspected (Aluze et al, 2002;Lee et al, 2000;Rosati et al, 2009) and also the inaccessibility of some inner surfaces for imaging. The curvature of the surface might give bands of illumination which might interfere with the image details and often gives an impression of false defects.…”

Section: Motivationmentioning

confidence: 99%

Identification of defects on highly reflective coated ring components using dark field illumination and image segmentation using simple thresholding technique

Singaperumal

Ramamoorthy

Sureshrao

et al. 2011

IJAAC

View full text Add to dashboard Cite

This paper deals with inspection of highly reflective chrome coated rings used in textile machinery using machine vision. These rings are mass produced in very large numbers in a textile machinery manufacturing company and the inspection was carried out manually using optical methods. Introduction of automated vision inspection had not helped the industry fully to achieve 100% quality inspection. Moreover, the whole procedure had to be made quicker in view of the number of components manufactured which is of the order of hundreds of thousands and to be inspected 100% thereafter. Also in order to improve inspection speed and to ensure 100% quality inspection, it was absolutely essential to improve the complete inspection process. Also a proper sorting algorithm was to be considered to classify defective and nondefective components. The effect of curvature, reflective nature of material and the real-time inspection make the imaging and defect detection and classification difficult. The use of dark field illumination has been tried out in this work to quicken the process of defect detection and also to improve the accuracy of detection. The images of the defects captured using a CCD camera

show abstract

Machine vision system for curved surface inspection

Cited by 21 publications

References 2 publications

A Machine Vision System for Robust Sorting of Herring Fractions

A Machine Vision System for Robust Sorting of Herring Fractions

Accurate 3D Reconstruction using a Turntable-based and Telecentric Vision

Identification of defects on highly reflective coated ring components using dark field illumination and image segmentation using simple thresholding technique

Contact Info

Product

Resources

About