Machine Vision Handbook (机器视觉手册)

1 Machine Vision for Industrial

Applications

Bruce G. Batchelor

Cardiff University, Cardiff, Wales, UK

1.1 Natural and Artificial Vision .................................................... 3

1.2 Artificial Vision . ................................................................. 5

1.3 Machine Vision Is Not Computer Vision . . . .. . . . .. . ............................10

1.3.1 Non-industrial Applications . . . . . . . . .. . . . .. . . . .. . . . .. . . . . . . . . . . . . . . . . . . . . . . .. . . . .. 15

1.3.2 Does It Matter What We Call Our Subject? . . . . .. . . . . . . . . . . . . . .. . . . . . . . .. . . . .. . . 19

1.5.1 Systems Engineering . .. . . . .. . . . . . . . . . . . . . . . . . . . . . . .. . . . .. . . . .. . . . .. . . . . . . . .. . . . . .. 25

1.6 Design Tools . . . .. ................................................................27

1.6.1 Image Acquisition . .. . . . .. . . . . . . . .. . . . .. . . . . . . . . .. . . . . . . . . .. . . . .. . . . . . . . .. . . . .. . . . . 28

1.6.2 Algorithm Design . . . . . . . . . . . . . .. . . . . . . . . .. . . . .. . . . . . . . . .. . . . .. . . . . . . . .. . . . .. . . . . . . 28

1.7 Overview of This Book .. . . . . . ...................................................31

1.7.1 General Principles .. . . . . . . . . . . . . . .. . . . . . . . .. . . . .. . . . .. . . . .. . . . .. . . . . . . . . .. . . . . . . . . . 32

1.7.1.1 Chapter 2: Inspecting Natural and Other Variable Objects . . . .. . . . .. . . . . . . . . . . . 32

1.7.1.2 Chapter 3: Human and Animal Vision . . . . . . . . .. . . . . . . . . . . . . .. . . . .. . . . . . . . . .. . . . 33

1.7.1.3 Chapter 4: Colour Vision .. . . . .. . . . . . . . . .. . . . . . . . . . . . . . .. . . . . . . . .. . . . .. . . . .. . . . .. . 34

1.7.1.4 Chapter 5: Light and Optics . . . . . . . . . .. . . . . . . . . .. . . . .. . . . .. . . . .. . . . .. . . . . . . . .. . . . . 34

1.7.1.11 Chapter 12: X-ray Inspection . . . . .. . . . .. . . . . . . . . .. . . . .. . . . . . . . .. . . . .. . . . . . . . . .. . . . 41

1.7.1.12 Chapter 13: Illumination-Invariant Image Processing . . . . . .. . . . . . . . . .. . . . .. . . . .. 42

1.7.1.13 Chapter 14: Basic Machine Vision Techniques . . . .. . . . .. . . . .. . . . .. . . . . . . . .. . . . . .. 42

Bruce G. Batchelor (ed.), Machine Vision Handbook, DOI 10.1007/978-1-84996-169-1_1,

#

Springer-Verlag London Limited 2012

1.7.1.14 Chapter 15: Imaging and Range Image Processing . . .. . . . . . . . . .. . . . .. . . . . . . . . . . . 43

1.7.1.15 Chapter 16: Colour Recognition . .. . . . . . . . .. . . . .. . . . .. . . . .. . . . .. . . . . . . . . .. . . . . . . . 44

1.7.1.16 Chapter 17: Algorithms, Approximations and Heuristics . . . . . . . .. . . . .. . . . .. . . . . 45

1.7.1.17 Chapter 18: Object Location Using the Hough Transform .. . . . . .. . . . .. . . . . . . . .. 46

1.7.1.23 Chapter 24: Pattern Recognition . . . . . . .. . . . . . . . . . . . . . .. . . . . . . . . .. . . . .. . . . . . . . .. . . 50

1.7.1.24 Chapter 25: Implementing Machine Vision Systems Using FPGAs . . . .. . . . .. . . . 51

1.7.2 Applications Case Studies and Practical Considerations . . . .. . . . .. . . . .. . . . . . . . . . 51

1.7.2.1 Chapter 26: Very Low-Cost In-Process Gauging System . . .. . . . .. . . . . . . . . . . . . . .. 52

1.7.2.2 Chapter 27: Automated Handling of Coils with Flying Leads . . .. . . . . . . . .. . . . .. 52

1.7.2.3 Chapter 28: A Telecentric Vision System for Broach Verification . . . . .. . . . .. . . . . 53

1.7.2.4 Chapter 29: Challenges of Low Angle Metal Surface (Crosshead)

Inspection .. . . . .. . . . .. . . . .. . . . . . . . . .. . . . . . . . . . . . . .. . . . .. . . . .. . . . .. . . . . . . . . .. . . . .. . . 53

1.7.2.5 Chapter 30: A Machine Vision System for Quality Grading of

Painted Slates . . . . . . .. . . . . . . . . . . . . .. . . . .. . . . .. . . . .. . . . . . . . . .. . . . .. . . . . . . . .. . . . .. . . . . 54

Abstract: Machine Vision is related to, but distinct from Computer Vision, Image Processing,

Artificial Intelligence & Pattern Recognition. The subject is concerned with the engineering of

integrated mechanical-optical-electronic-software systems for examining natural objects and

materials, human artifacts and manufacturing processes, in order to detect defects and improve

IR and X-ray sensor arrays, as well as laser scanners), digital, analogue and video electronics,

practices and quality assurance methods. Machine Vision is not a scientific endeavour; it is a

branch of Systems Engineering. Hence, consideration of application requirements pays a key

role in the design of practical vision systems. The basic philosophy of Machine Vision is set out

in this chapter and the structure of this book is outlined. This is a pragmatic, empirical,

experimental subject and is not unduly burdened by the search for mathematical rigour, or

theoretical purity. There is one simple design maxim: if it works, use it! This is justified here by

consideration of lessons learned from four practical application studies. Later in the book these

same ideas will emerge many times over.

Vision is critical for the survival of almost all species of the higher animals, including fish,

amphibians, reptiles, birds and mammals. In addition, many lower animal phyla, including

insects, arachnids, crustacea and molluscs possess well-developed optical sensors for locating

food, shelter, a mate, or a potential predator. Even some unicellular organisms are photosen-

sitive, although they do not have organs that can form high resolution images. Vision bestows

great advantages on an organism. Looking for food during foraging, or prior to giving chase, is

very efficient in terms of the energy expended. Animals that look into a crevice in a rock, to

in both the forest and a modern industrial society would be impossible without the ability to

see. Human society is organised around the highly refined techniques that people have

developed to communicate visually. People dress in a way that signals their mood, social

standing and sexual availability. Commerce is dependent upon hand-written, printed, and

electronic documents that all convey messages visually. Education relies heavily upon the

students’ ability to absorb information that is presented visually. Writers of technical books,

such as this, exploit the reader’s ability to understand complex abstract ideas through the visual

medium provided in printed text, diagrams and pictures. The leisure and entertainment

industries rely on the fact that vision is the dominant human sense. So important is vision

for our everyday lives that its loss is regarded by most people as one of the worst fates that could

Manufacturing industry is critically dependent upon human beings’ ability to analyse

complex visual scenes. This is exploited in innumerable and highly variable tasks, ranging

from the initial design process, through initial forming, machin ing , finishing, assembly,

and quality control, to final packing. The value of vision as a safe way to sense many of the

about a wide variety of features of an object, including its shape, colour, texture, surface

finish, coating, surface contamination, etc. Vision is extremely versatile and we can employ

a wide range of optical techniques to widen its range of application even further. For

example, it is possible to use stroboscopic light sources, structured ligh ting, colour filters,

coherent illumination polarise d light, fluorescence and many other clever optical ‘‘tricks’’

to sense a very broad range of physical prop erties.

Human vision is truly remarkable in its ability to cope with both subtlety and huge

variations in intensity, colour, form and texture. The dynamic range is enormous: about

10

matching pictures w ith great precision. A person can drive a car at high speed, yet a watch-

maker can make very fine adjustments to a mechanism. A player can strike the cue ball in the

game of billiards, or pool, with exquisite accuracy. A painter can capture the subtlety of colour

in a woodland scene on his canvas, while an experienced foundr y worker can identify the

temperature of an incandescent ingot from its colour very accurately. All of these visual skills

are exercised for inspection, grading, sorting, counting and a great many other tasks required

for manufacturing industry.

Despite its obvious benefit to industry, human vision has its limitations (

factories must not be exposed to high temperatures, toxic/stifling atmospheres, high levels of

smoke, dust and fumes, biological hazards, risk of explosion, excessively high noise levels, and

ionizing radiation. On the other hand, people can damage delicate products, by clumsy

handling and readily infect biological materials, by coughing, sneezing, shedding skin and

hair. Furthermore, people cannot always apply objective inspection criteria rigorously, partic-

ularly when aesthetic properties of highly variable items are to be judged. For these reasons,

engineers have long dreamed of building a machine that can ‘‘see.’’ Such a machine should be

able to sense its environment optically, using a video camera, or other imaging sensor,

interfaced to an electronic information-processing system: a standard computer, or dedicated

electronic hardware.

attempt to emulate human, or animal, vision systems. Indeed, the requirement that industrial

vision systems must be fast, cost effective and reliable, together with our rather limited

knowledge about how people and animals actually see, make this approach unprofitable. By

studying vision in nature, we may gain insight but we must be flexible and pragmatic when

designing industrial vision systems.

When the author of this chapter is asked by a non-technical person what he does for

a living, he replies that, as an academic, he studies machines that consist of a television (more

properly, video) camera connected to a computer and uses them to inspect objects as they are being

made in a factory. Such a definition is acceptable around the dinner-table but not for a technical

book, such as this (

● Illumination