VERISENS APPLICATION DESCRIPTION: ORIFICE MEASUREMENT REQUIREMENTS A major manufacturer of plastic orifices needs to verify that the orifice is within the correct measurement band. Parts are presented to the VeriSens vision sensor and the sensor measures the area (corresponding to flow) of the plastic orifice. FIGURE 1: PART Figure 1 is a picture of the part. It is approximately 1 7/8 in diameter, approximately ¼ thick with an orifice in the center that should be approximately.0030 in diameter. APPLICATION CONSIDERATIONS RESOLUTION/ MEASUREMENT ACCURACY The inspections were performed with a VeriSens with a 1.5 Megapixel sensor. Pixel resolution is 1280 x 960. We make a distinction between a measurement application and one of gauging. As an on-line measurement application, cost considerations led to the use of a standard C-mount lens as opposed to microscope-type (cost factor approximately a factor of 12x for the microscope) optics, which would certainly provide more than adequate magnification. The Field of View (FOV) as tested is approximately 1¼. Therefore, the pixel resolution is approximately.00125 with equivalent measurement accuracy (regarding accuracy: a telecentric lens would improve accuracy ay an approximate 3x cost increase). Page 1 of 9
LIGHTING Figure 2 illustrates the lighting technique used in this application, back lighting. That is, the part is located between the light source and the sensor so as to present a silhouette. In this case, light shines through the orifice allowing its measurement (See Figure 3). In actual production, the parts would be located in open nests with a red LED backlight located beneath them. OPTICS This preliminary test indicates that standard optics should be adequate to perform this inspection. A C-mount lens for this application is in the $400 FIGURE 2: BACKLIGHTING range. A microscope lens, which would improve accuracy, is in the $5000 range. A telecentric lens, which would improve accuracy, is in the $1200 range. We suggest that a sample set of parts be run using a standard lens. If the results prove to be inadequate, then other lens options could be considered. INSPECTION METHODOLOGY Figures 4 through 7 illustrate the inspection methodology. It is fairly straightforward consisting of only one software operations. After the image is acquired, a trigger prompts the VeriSens to acquire an image and the Area Tool is used to measure the area of the orifice. The Figures illustrate the measurement executions, shown in the Image Window with the green tools indicating that a previously taught value was successfully obtained. Note the measurement values in the Results Window. These values are in pixels, but could be transformed into millimeters if desired. FIGURE 3: TEST SETUP Page 2 of 9
FIGURE 4: PART 1,.00131 PASS Figures 4 and 6 illustrate what happens conforming part is introduced to the VeriSens sensor. Note the Results Window. Page 3 of 9
FIGURE 5: PART 2,.00175 FAIL Figures 5 and 7 illustrate non-conforming parts. Again, note the Results Window. Page 4 of 9
FIGURE 6: PART 3,.00129 PASS Page 5 of 9
FIGURE 7: PART 4,.00203 FAIL. CONCLUSION This inspection can be performed successfully provided the following criteria are met: 1) Parts must be backlit. 2) Orientation of the camera relative to the orifice is important 3) Additional considerations are a red LED and possibly an upgraded lens. Page 6 of 9
CONTACT Baumer,Ltd. USA Pete Kepf, Field Application Engineer pkepf@baumer.com 865-518-1985 Page 7 of 9
GLOSSARY Back Lighting Lighting technique in which the part is positioned between the light source and the camera resulting in a silhouette Contrast Measured in grayscale (Black = 0; white = 255). A minimum contrast must be present in order to differentiate a feature from its background. Feature Extraction The ability of machine vision software to separate a feature from its background and determine some attribute (area, brightness, etc.) Camera Front (or Top) Lighting Lighting technique in which the light source and the camera are on the same side of the part to be inspected. Light Source Part Optical Character Page 8 of 9
Recognition (OCR) The ability of machine vision software to determine the value of a string of previously unknown characters. Optical Character Verification (OCV) The ability of machine vision software to determine whether or not the value of a string of previously taught characters matches the current image. Resolution Measured in Units of measure per pixel. Field of View minimum number of pixels on the sensor. For example, a 1 FOV with a VGA sensor has a resolution of approximately.002 Resolution = FOV/pixels Resolution = 1 /480 Resolution =.002 / pixel Telecentric Lens Telecentric Lenses are used in imaging systems to make objects appear to be the same size independent of their location in space. Telecentric Lenses remove the perspective or parallax error that makes closer objects appear to be larger than objects farther from the lens, increasing image quality compared to conventional lenses. Telecentric Lenses are ideal for use in a variety of applications, including metrology, gauging, CCD based measurement, or microlithography. Page 9 of 9