Industrial Image Processing: Visual Quality Control in Manufacturing

ISBN-13: 9783642636424 / Angielski / Miękka / 2012 / 353 str.

Christian Demant; Bernd Streicher-Abel; Peter Waszkewitz

Opis

Digital image processing has become a key technology in the area of manu facturing and quality control. Increasing quality demands require inspection of every single part, which in turn will lead to a much more widespread use of automatic visual inspection systems in the near future. Furthermore, the documentation requirements of ISO 9000 and similar quality control standards can only be met by fully automated, networked inspection systems. On the other hand, despite a multitude of successful applications, digital image processing has not yet established itself as an accepted element of man ufacturing technology. This holds true for the industrial practice as well as for the training of engineers. Digital image processing is still widely regarded as some kind of secret lore, mastered only by a small number of expensive -- experts. This impression of incomprehensibility frequently leads to the accusation of unreliability. The manufacturers of digital image processing systems in the industry are not least responsible for this state of affairs, due to their policy of giving the customer as little information as possible about the methods and technology used to inspect his products."

Spis treści

1. Introduction.- 1.1 Why write another book about image processing?.- 1.2 Possibilities and limitations.- 1.3 Types of inspection tasks.- 1.4 Structure of image processing systems.- 1.4.1 Hardware setup.- 1.4.2 Signal flow in the process environment.- 1.4.3 Signal flow within an image processing system.- 1.5 Solution approach.- 1.6 Introductory example.- 1.6.1 Character recognition.- 1.6.2 Thread depth.- 1.6.3 Presence verification.- 1.7 From here.- 2. Overview: Image Preprocessing.- 2.1 Gray scale transformations.- 2.1.1 Look-up tables.- 2.1.2 Linear gray level scaling.- 2.1.3 Contrast enhancement.- 2.1.4 Histogram equalization.- 2.1.5 Local contrast enhancement.- 2.2 Image arithmetic.- 2.2.1 Image addition and averaging.- 2.2.2 Image subtraction.- 2.2.3 Minimum and maximum of two images.- 2.2.4 Shading correction.- 2.3 Linear filters.- 2.3.1 Local operations and neighborhoods.- 2.3.2 Principle of linear filters.- 2.3.3 Smoothing filter.- 2.3.4 Edge filters.- 2.4 Median filter.- 2.5 Morphological filters.- 2.6 Other non-linear filters.- 2.7 Global operations.- 2.8 Key terms.- 3. Positioning.- 3.1 Position of an individual object.- 3.1.1 Positioning using the entire object.- 3.1.2 Positioning using an edge.- 3.2 Orientation of an individual object.- 3.2.1 Orientation computation using principal axis.- 3.2.2 Distance-versus-angle signature.- 3.3 Robot positioning.- 3.3.1 Setting of tasks.- 3.3.2 Image processing components.- 3.3.3 Position determination on one object.- 3.3.4 Orientation of an object configuration.- 3.3.5 Comments concerning position adjustment.- 3.4 Key terms.- 4. Overview: Segmentation.- 4.1 Regions of interest.- 4.1.1 Regions and objects.- 4.2 Thresholding.- 4.2.1 Thresholds.- 4.2.2 Threshold determination from histogram analysis.- 4.2.3 Gray level histograms.- 4.2.4 Generalizations of thresholding.- 4.3 Contour tracing.- 4.3.1 Pixel connectedness.- 4.3.2 Generating object contours.- 4.3.3 Contour representation.- 4.4 Edge based methods.- 4.4.1 Edge probing in industrial image scenes.- 4.4.2 Edge detection with subpixel accuracy.- 4.5 Template matching.- 4.5.1 Basic operation.- 4.5.2 Optimizing template matching.- 4.5.3 Comments on template matching.- 4.6 Key terms.- 5. Mark Identification.- 5.1 Bar code identification.- 5.1.1 Principle of gray-level-based bar code identification.- 5.1.2 Bar code symbologies.- 5.1.3 Examples of industrial bar code identification.- 5.1.4 Further information.- 5.2 Character recognition.- 5.2.1 Laser-etched characters on an IC.- 5.2.2 Basic configuration of the character recognition.- 5.2.3 Fundamental structure of a classifier application.- 5.2.4 Position adjustment on the IC.- 5.2.5 Improving character quality.- 5.2.6 Optimization in operation.- 5.3 Recognition of pin-marked digits on metal.- 5.3.1 Illumination.- 5.3.2 Preprocessing.- 5.3.3 Segmentation and classification.- 5.4 Block codes on rolls of film.- 5.5 Print quality inspection.- 5.5.1 Procedure.- 5.5.2 Print quality inspection in individual regions.- 5.5.3 Print quality inspection with automatic subdivision.- 5.6 Key terms.- 6. Overview: Classification.- 6.1 What is classification?.- 6.2 Classification as function approximation.- 6.2.1 Machine learning.- 6.2.2 Statistical foundations.- 6.2.3 Constructing classifiers.- 6.3 Instance-based classifiers.- 6.3.1 Nearest neighbor classifier.- 6.3.2 RCE networks.- 6.3.3 Radial basis functions.- 6.3.4 Vector quantization.- 6.3.5 Template matching.- 6.3.6 Remarks on instance-based classifiers.- 6.4 Function-based classifiers.- 6.4.1 Polynomial classifier.- 6.4.2 Multilayer perceptron-type neural networks.- 6.4.3 Representation of other classifiers as neural networks.- 6.5 Remarks on the application of neural networks.- 6.5.1 Composition of the training set.- 6.5.2 Feature scaling.- 6.5.3 Rejection.- 6.5.4 Arbitrariness.- 6.6 Key terms.- 7. Dimensional Checking.- 7.1 Gauging tasks.- 7.2 Simple gauging.- 7.2.1 Center point distance.- 7.2.2 Contour distances.- 7.2.3 Angle measurements.- 7.3 Shape checking on a punched part.- 7.3.1 Inspection task.- 7.3.2 Modeling contours by lines.- 7.3.3 Measuring the contour angle.- 7.4 Angle gauging on toothed belt.- 7.4.1 Illumination setup.- 7.4.2 Edge creation.- 7.5 Shape checking on injection-molded part.- 7.5.1 Computing radii.- 7.5.2 Remarks on model circle computation.- 7.6 High accuracy gauging on thread flange.- 7.6.1 Illumination and image capture.- 7.6.2 Subpixel-accurate gauging of the thread depth.- 7.7 Calibration.- 7.7.1 Calibration mode.- 7.7.2 Inspection-related calibration.- 7.8 Key terms.- 8. Overview: Image Acquisition and Illumination.- 8.1 Solid-state sensors.- 8.1.1 CCD sensor operation.- 8.1.2 Properties of CCD sensors.- 8.1.3 Image degradation.- 8.2 Standard video cameras.- 8.2.1 Basic structure.- 8.2.2 The video standard.- 8.2.3 Sampling of the line signal.- 8.2.4 Extensions of the video standard.- 8.2.5 Image quality.- 8.3 Other camera types.- 8.3.1 Progressive scan cameras.- 8.3.2 Asynchronous cameras.- 8.3.3 Digital cameras.- 8.3.4 Line-scan cameras.- 8.3.5 Additional camera properties.- 8.4 Transmission to the computer.- 8.4.1 Basic operation of a frame grabber.- 8.4.2 Frame grabbers for standard video cameras.- 8.4.3 Frame grabbers for other camera types.- 8.4.4 Direct digital transmission.- 8.5 Optical foundations.- 8.5.1 F-number.- 8.5.2 Thin lens imaging equation.- 8.5.3 Depth of field.- 8.5.4 Typical imaging situations.- 8.5.5 Aberrations.- 8.5.6 Lens selection.- 8.5.7 Special optical devices.- 8.6 Illumination technology.- 8.6.1 Light sources.- 8.6.2 Front lighting.- 8.6.3 Back Lighting.- 8.7 Key terms.- 9. Presence Verification.- 9.1 Simple presence verification.- 9.1.1 Part geometry.- 9.1.2 Illumination.- 9.1.3 Position adjustment.- 9.1.4 Segmentation.- 9.1.5 Evaluation.- 9.1.6 Segmentation with template matching.- 9.2 Simple gauging for assembly verification.- 9.2.1 Illumination.- 9.2.2 Inspection criteria.- 9.2.3 Object creation and measurement computation.- 9.2.4 Position adjustment.- 9.3 Presence verification using classifiers.- 9.3.1 Illumination.- 9.3.2 Inspection of the caulking.- 9.3.3 Type verification of the flange.- 9.4 Contrast-free presence verification.- 9.5 Key terms.- 10. Overview: Object Features.- 10.1 Basic geometrical features.- 10.1.1 Enclosing rectangle.- 10.1.2 Area and perimeter.- 10.1.3 Center of gravity.- 10.1.4 Axes and radii.- 10.2 Shape-descriptors.- 10.2.1 Curvature.- 10.2.2 Fiber features.- 10.2.3 Euler’s number.- 10.2.4 Moments and Fourier descriptors.- 10.3 Gray level features.- 10.3.1 First-order statistics.- 10.3.2 Texture features.- 10.4 Key terms.- 11. Outlook: Visual Inspection Projects.- A. Mathematical Notes.- A.1 Backpropagation training.- A.1.1 Neural networks — concept and history.- A.1.2 Fundamentals.- A.1.3 Backpropagation.- A.2 Computation of the depth of field.- A.2.1 Limit distances.- A.2.2 Depth of field at infinite distance.- A.2.3 Dependence of the depth of field on the focal length.- B. The Companion CD.- References.

Blurb

This practical introduction focuses on how to build integrated solutions to industrial vision problems from individual algorithms. It gives a hands-on guide for setting up automated visual inspection systems using real-world examples and the NeuroCheck(Registered Trademark) software package. This software has actually been tested on a production line. Based on many years of experience in industries, the editors explain all the (mostly unpublished but essential) details encountered in the creation of real-world vision systems. With the original NeuroCheck (R) software package and all the example images included on CD-ROM, readers can work their way through the described inspection tasks and carry out their own experiments.