Understanding And Applying Machine Vision Part 13 ppsx

Tessellation Pixel pattern. Most tesselations are square (square pixels); some are rectangular (rectangular pixels). Texture Local variation in pixel values that repeats in a regular or random way across a portion of an image or object. Texture is concerned with the spatial distribution of the gray shades and discrete tonal features. When a small area of the image has little variation of discrete tonal features, the dominant property of that area is gray shade. When a small area has wide variation of discrete tonal features, the dominant property of that area is texture. There are three things crucial in this distinction: (1) the size of the small areas, (2) the relative sizes of the discrete tonal features, and (3) the number of distinguishable discrete tonal features. Texture Gradient Refers to the increasing fineness of visual texture with depth observed when viewing a two-dimensional image of a three-dimensional scene containing approximately uniform textured surfaces. Three-Dimensional Analysis Type of vision algorithm that develops a three-dimensional model of a part and matches it to stored models. Page 378 Threshold Intensity (specific pixel value) below which a stimulus produces no effect or response. Dynamic threshold: When threshold result depends on the location of a pixel, the characteristic pixel value and some local property. Global threshold: When threshold result depends only on comparison of a pixel value with a global constant (a single brightness value, a single color value, etc.). Local threshold: When threshold result depends on comparison of a joint local property and a characteristic pixel value with a single decision value, e.g., Threshold is based on comparing the difference between pixel value and neighborhood value with a global value. Thresholding Scene segmentation process based on converting a gray scale image into a binary image by reassigning pixel gray levels to only two values. Regions of an image are separated based on pixel values above and below a chosen intensity level. Threshold Value Decision point above or below which a decision is made, e.g., a pass or fail point. Throughput Rate Generally refers to the number of objects to be examined per unit of time. Time of Flight Technique of inferring shape from reflection of active light signal off object based on measuring elapsed time for signal to return when an impulse source is used or by modulating a CW source signal and matching the return signal to measure phase differences that are in turn interpreted as range measurements. Token In a data flow machine, term associated with combination of data word and label or tag. Tolerance Amount that establishes the range upon which to base the differentiation between good and bad products. Top-Down Approach Goal-directed image analysis approach in which the interpretation stage is guided in its analysis by trial or test descriptions of a scene. Sometimes referred to as "hypothesize and test." Tracking Processing sequences of images in real time to derive a description of the motion of one or more objects in a scene. Transforms Various mathematical transformations applied to image data for the purpose of analysis. See Image Transformation. Transition In a binary image, the point where the pixels change between light and dark. Transition Counting Modeling of a scene based on the number of white-to black and black-to-white pixel changes. Translation Movement left or right, up or down, but not rotated; geometric operation that shifts the position of an image from its original position. Transmission Passage of light or other signal. Transmittance, Transmittance Coefficient Ratio of the energy per unit time per unit area (radiant power density) transmitted through the object to the energy per unit time per unit area incident on the object. In general, transmittance Page 379 is a function of the incident angle of the energy, viewing angle of the sensor, spectral wavelength and bandwidth, and the nature of the object. Tree Hierarchical representation of a scene. Triangulation Method of determining distance by trigonometry. Two and One-Half Dimensions (2 1/2 D) Photometric stereo; see Needle Diagram. U Ultraviolet (UV) Region of the electromagnetic spectrum adjacent to the visible portion of the spectrum but with wavelengths between 100 and 400 nanometers. Union (of Two Images) Logical operation forming a new image that is black at all points where either of the two images are black. V Vector Encoding Method of characterizing line segments extracted from a scene, specifying each segment as a pair of image coordinates corresponding to the segment's end points. Verification Activity providing qualitative assurance that a fabrication or assembly process was successfully completed. Vertex Point on a polyhedron common to three or more sides. Vertical Resolution Number of horizontal lines that can be seen in the reproduced image. Vertical Sync Circuit to retrace the scan from bottom to top. Video Analog time-varying output signal from an image sensor that conveys the image data. Video Image Image in electronic signal format capable of being displayed on a cathode-ray tube screen. The video signal is generated from such devices as a vidicon or flying spot scanner, which convert an image from photographic form to video signal form by scanning it line by line. The video signal itself is a sequence of signals, the signal representing the line of the scanned image. Vidicon Image pickup tube in which a change density pattern is formed by photoconduction and stored on that surface of the photoconductor that is scanned by an electron beam. Visible Light that can be seen by the eye. Having wavelength between 400 and 750 nanometers. Vision Process of understanding the environment based on sensing the light level or reflectance property of objects. Von Neumann Architecture Current standard computer architecture that uses sequential processing. W Walsh-Hadamard Transform See Hadamard Transform. Wavelength Reciprocal of frequency. The distance covered by one cycle or event. Window Selected portion of an image. Also, a limited range of gray scale values. Page 380 Windowing Technique for reducing data-processing requirements by electronically defining only a small portion of the image to be analyzed. All other parts of the image are ignored. Wireframe Model Three-dimensional model, similar to a wireframe, in which the object is defined in terms of edges and vertices. Z Zoom To enlarge or reduce the size of an image. It may be done electronically or optically. Zoom Lens Optical system of continuously variable focal length, the focal plane remaining in a fixed position. Page 381 Appendix B— Machine Vision Application Checklist This is a form to assist in developing ideas and requirements for machine vision applications. It contains many typical questions to help determine feasibility, benefits and cost. Permission is granted to use this form. Section 1— Production Process 1. What do you make? 2. How do you make it? 3. What is the expected life of the product? 4. Is the product and the problem going to be around long enough to justify the purchase of a system? 5. Why do you need to inspect or control the process? Are there problems? Is process improvement the goal? 6. What is the current reject rate of bad parts? 7. What is the accuracy of the current inspection system? Page 382 8. If you inspect or improve control, what are the specific benefits you expect to achieve? 9. Will the system be for a new line or an old line (retrofit)? 10. Does your application involve: one object at a time? Multiple objects? How many different objects? What are the different part numbers? 11. Is it a batch operation? or a continuous dedicated process/line? 12. What are the changeover times and the frequency of changeovers? 13. What are the skill levels involved in changeover? 14. How is inspection and/or function to be replaced currently being performed? Is it effective? 15. Is inspection to be: On line? or Off line? 16. Must every produced item be inspected, or can you randomly sample? 17. Will new part models or variations be added to the system at a later date? Define any potential future inspections that may be required of the same machine vision systems: 18. Are product design or production process changes anticipated? 19. Where do parts come from? 20. Can rejected parts be repaired? 21. Can vision assist in the diagnosis? 22. Where do pass and fail parts go? 23. When is the machine vision system needed by? 24. How many shifts will the system be used? 25. How many lines/machine will the vision system be needed for? 26. What is the attitude of the plant floor people towards machine vision/automation? 27. What is the attitude of the plant's management toward machine vision/automation? 28. What is your attitude towards machine vision/automation? 29. Can a representative sample of parts be provided to system vendors or integrators for evaluation? Page 383 30. Can drawings be provided? 31. Can video of line be provided? 32. Can the vision suppliers observe production at your facility? Section 2— Benefits of Inspection 1. When an incorrect or flawed part escapes detection, what are the downstream effects? (quality, repair, machine downtime, etc.): 2. If a bad part is assembled, does it cause problems with the overall assembly? 3. If inspection is implemented, can any downstream testing requirements be relaxed? 4. If inspection is implemented, do you expect the yield through test to be improved? 5. At the inspection point, what is the cost (qualitative or quantitative) of a bad (faulty) part that escapes? 6. At the inspection point, what is the cost (qualitative or quantitative) of a good part that is falsely rejected? Section 3— Application 1. Describe the application: 2. What distinguishes a bad part from a good part? 3. Generically, does the application involve: Gaging? (Show a sketch or drawing, if possible. Highlight critical items.) What are tightest tolerances? On what specific dimension? What is the design goal for accuracy? What is the design goal for repeatability? Are there features that serve as references? Describe calibration requirements: Page 384 Assembly Verification? Dimensions of assembly: Presence/absence: Orientation: What is the smallest piece to be verified? What are the dimensions of that piece? What is the largest piece to be verified? What are the dimensions of that piece? Do you also need to verify the correctness of the part? Flaw inspection? Describe flaw types: What is the smallest size flaw? Does the flaw affect surface geometry? Does the flaw affect surface reflectivity? Is it more of a stain? Is classification of flaws required? Location analysis? What is the design goal for accuracy? What is the design goal for repeatability? What is the area over which "find" is required? Describe calibration requirements: Pattern recognition? What is the size of the pattern? Describe the differences between patterns: Is there a background pattern? Color? Geometric? Number of different patterns? Purpose: to identify? to sort? other? Page 385 Are pattern differences geometric? Are pattern differences photometric? Are pattern differences color? Is application specifically OCR? OCV Handwritten characters? Fixed font? Variable font? What is the height of the characters? What is the stroke width? What is the spacing between characters? What is the spacing around the characters? How many characters in a string? How many lines? Describe background What is the color of the characters? What is the color of the background? Section 4— Part to Be Inspected 1. Describe the part(s) to be inspected (consider those conditions that can change appearance of part or background). Are drawings available? 2. What is specific material (steel, plastic, etc.)? 3. What is specific finish (texture) like? Is the surface finish the same on all faces of the part? Is the surface finish the same for all part numbers and/or production runs? Describe any differences: Describe the platings: Coating? Thin films (oils, mist, etc.)? Paint? Dull? Glossy? Page 386 Specular? Highly reflective (mirror like)? Poorly reflective? Dull? Matte? Will the reflectivity of the part change from part to part? Over time? 4. Are there any machining marks on the part? Does the part generally have scratches, nicks, burrs, dents, etc.? Is there any porosity on the parts? 5. What are the object's shapes? —Flat? Curved? Gently curved? Other? Irregular? Grooved surface? Sharp radii (prismatic)? Mixed geometric properties? 6. Is part always oriented in the same direction? 7. What is the temperature of the part? 8. What is the size of part(s)? Smallest: Largest: 9. Are there different colors for different models? Does the color of the part change from part to part? Color: - single hue - variations in saturation - subtle color variations - discrete color variations - mixed with broad and discrete colors 10. Discuss conditions such as warpage, shrinkage, bending that could be experienced. 11. Is there any change in appearance over time due to environment? (rust inhibitors, corrosion, lubricants, dirt, perishability, etc.) 12. Are there any markings on the part? 13. Is it possible to make a refernce mark on the part, if necessary? 14. What are part appearance variables? Page 387 15. Is surface translucent? Describe variations in translucent optical density/degree of opaqueness: 16. Is surface transparent? Totally? Partially? 17. Is part sensitive to heat? 18. Is object sensitive to light? If yes, what type of light? Ultraviolet? Visible? Infrared? Section 5— Material Handling? 1. Describe the material handling system (current or planned): 2. Is object subject to damage in handling? Describe precautions to take in handling: 3. Will inspection be done at a station that also performs other functions? Describe those functions: 4. What is the production rate? (How many parts per minute on average?) During production catch up mode or peak rates? 5. Are there any expected changes that might affect the above rates? 6. Are parts static/indexed? or moving continuously? If indexed: How long stationary? Total in-dwell-out time: What is the settling time? Is there any acceleration and, if so, where in the cycle? If parts are moving continuously, what is the speed? Regulation of that speed? Page 388 7. What are the maximum positional variations that can be expected? +/-X +/-Y +/-Z +/- degrees around X direction +/- degrees around Y direction +/- degrees around Z direction 8. How much spacing is there between parts? Is the spacing random? or constant? Part spacing repeatability: Do parts ever touch? Do parts ever overlap? 9. Is there more than one stable state involved? How many? 10. If there will be multiple inspections, will the part maintain the same orientation throughout the process? 11. What is the volume envelope available for an inspection station? 12. Are there any restrictions or obstructions to viewing the product? 13. Is there a weight constraint? 14. How close can an associated electronic enclosure be located? 15. How far will the vision system controller be from other equipment that the system will be interfaced with? 16. Describe any other physical constraints surrounding the proposed installation site: 17. If conveyor - what type? What color? What are the appearance variables of the conveyor (specular, uniformity, over time, etc.)? 18. Is a by-pass mode required? 19. Describe action to take when reject is detected: [...]... Bin-picking, 13 Biometrics, 22 Bit, 50 Blobs, 195 Boolean analysis Brightness, 75 scaling, 167 C CAD/CAM, 4, 5 Cameras, 133 –151, 283 (see also Image; Image scanning; Television cameras) CCTV, 129, 133 EIA RS-170 standards, 133 , 134 , 136 features, 137 138 asynchronous reset, 135 automatic black level, 138 automatic gain control, 138 automatic lens drive circuit, 138 automatic light range, 138 electronic... control, 138 automatic lens drive circuit, 138 automatic light range, 138 electronic shuttering, 135 gamma correction, 138 minimum illumination, 138 resolution, 138 field, 133 frames, 133 , 135 genlock, 136 interlaced progressive, 134 , 137 raster scanning smear, 134 solid state, 9 sync, 134 timing, 136 video signal, 135 Captive system integrator, 33 Calibration, 204 Classification, 166 Closing, 185 Coding,... Test/Buyoff Procedure 1 How will we be sure the machine is functioning properly? (Describe performance test hurdles.) 2 Tests at vendor site: a Can good, bad, and/ or marginal parts be provided? b What is the sample-size for each challenge? c Define acceptability criteria for each challenge d Define part variations or parts to be used during acceptance testing: e Define part position variations to be used during... acceptance testing: g Define environmental conditions: 3 Tests at installation site: a Can good, bad, and/ or marginal parts be provided? b What is the sample-size for each challenge? c Define acceptability criteria for each challenge d Define part variations or parts to be used during acceptance testing: e Define part position variations to be used during acceptance testing: f Define lighting variations to... device? If statistics are required, how often will the reports be generated? Will the reports need to be printed and/ or displayed? 13 What false reject rate is acceptable? 14 What escape rate or false acceptance rate is acceptable? Page 390 Section 7— Machine Interfaces 1 Alarms desired? 2 What other machines must this system be integrated with mechanically or electrically? 3 What event will trigger an inspection?... an inspection? How will the event be detected? How will this be communicated to the inspection system? 4 How will the results of the inspection be communicated and implemented? 5 Describe machine interfaces required/handshaking signals, etc.: part in position sensor type opto-isolation AC DC voltage level RS 232 signal conditioning required RS 422 Parallel PLC Ethernet RS 449 IEEE 488 PCI MAP Other... Describe the air quality: Dust/Smoke Steam Oil 3 Ambient light (type-incandescent, fluorescent, etc.): 4 Dirt on parts: Lubricant on parts: 5 Wash-down requirements: 6 Corrosive atmosphere: 7 Temperature range: Operating 8 Humidity range: Operating Storage Storage 9 Radiation EMI RFI 10 Shock: 11 Vibration: 12 Hazardous environment: 13 Utilities available: Compressed air How clean Water Input power -... Shipping: 3 Installation: 4 Warranty: 5 Spare parts: 6 Cost: 7 Documentation (instruction manuals for operator, maintenance, engineering, programming, prints, schematics, spare parts lists, software, etc.): 8 Training - Where and when? Details (subjects to be covered): operator: maintenance: programming: For how many? 9 Software Issues - ownership of Revisions Will vendor be required to support? For... 244–250 Erosion, 185 Escape rate, 321 F False reject rate, 321 Feature extraction, 166, 192–204 (see also Coding) Fiber optics, 90, 100, 345 Field, 133 , 135 Filters, 112 Fingerprint, 207 Flying spot scanner, 24, 140 Food industry applications, 265–267 Frame, 133 , 135 Frame addition, 160 (see also Smoothing) Frame buffer, 158 Frame grabber, 9, 48, 158 Frame store, 75 Frame subtraction, 160 Fourier transform,... shades; Shades of gray) Gray scale system, 53, 203, 214, 291 Page 398 Gray shades (see also Gray levels; Shades of gray), 50 H Histogram, 5, 76, 195, 205 Hue, 50, 83, 84, 298 Human vision, 86, 298 people, 41 vs machine vision, 19, 40, 41, 45, 46, 75 I Iconic, 216 Identifying applications, 70 Illumination optics (see also Light) collimated, 103 condenser lens, 101 dark field, 108 diffusers, 102 fiber . cameras) CCTV, 129, 133 EIA RS-170 standards, 133 , 134 , 136 features, 137 138 asynchronous reset, 135 automatic black level, 138 automatic gain control, 138 automatic lens drive circuit, 138 automatic. light range, 138 electronic shuttering, 135 gamma correction, 138 minimum illumination, 138 resolution, 138 field, 133 frames, 133 , 135 genlock, 136 interlaced progressive, 134 , 137 raster. anticipated? 19. Where do parts come from? 20. Can rejected parts be repaired? 21. Can vision assist in the diagnosis? 22. Where do pass and fail parts go? 23. When is the machine vision system needed