Surface Inspection

Scratch detected on laptop housing using contrast measurement.
Customers of today’s consumer electronics devices expect a flawless product out of the box. Scratched, dented, and otherwise imperfect products result in costly returns and may damage a brand’s reputation and future business. Quality control operations that are intended to prevent surface defects rely largely on human inspectors to apply their acute perception and immediate judgment of product quality during production to ensure defective products are not shipped to consumers. However, the faster the production rate, more complex the product, or more obscure the defect, the more difficult it is for humans to keep up with throughput demands while adhering to the need for absolute quality.
Radiant ProMetric® Imaging Photometers are engineered for the measurement and detection of subtle variations in light uniformity across illuminated components. Based on these photometric measurement capabilities, Radiant imagers are ideally-suited for detecting anomalous features on surfaces and grading these defects based on their location, scope, and severity. Radiant systems also offer all of the benefits of automated inspection, including objective analysis and quantifiable results, enabling error tracking for automation and operational improvement. An extensive set of machine vision and photometric evaluation tools are also available for production-level monitoring and pass/fail test sequencing using Radiant’s TrueTest™ Automated Visual Inspection Software.

Unmatched Camera Resolution and Dynamic Range

The performance of a machine vision inspection system is dependent on the acquisition of high-quality images, enabling the system to “see” defects in greater detail and apply meaningful evaluations. The resolution of a standard machine vision camera sensor will typically range from one to five megapixels. Leveraging the superior resolution (up to 29MP) and dynamic range (up to 73.4 dB) of Radiant’s ProMetric Imaging Photometers, Radiant machine vision systems capture unmatched image clarity. These systems apply their high dynamic range to detect hairline variations on part surfaces caused by reflections of light (like shadows that indicate a scratch), while high resolution CCD sensors allow the systems to image and classify defects with extreme precision. Radiant cameras can detect defects unnoticed by human inspectors, including light scratches on glass and foreign particles.

Unknown Defects in Unknown Locations

Dent detected on tablet back.

Radiant offers complete CCD-based camera systems that utilize advanced imaging technology and sophisticated algorithms to locate and characterize randomly-occurring cosmetic defects on device surfaces, including unknown defects in unknown or unpredictable locations on a part. By taking a wide area image of a device, Radiant systems can perform uniformity checks to detect hairline scratches, subtle dents, or even fingerprints on device surfaces. The systems can classify the severity of each defect against acceptable tolerances and apply pass/fail values, just as a human would use their judgment to determine a reasonable margin of error. Unlike human inspectors, however, Radiant solutions can apply numeric values to each defect and quantify these by size, shape, location, frequency, and rate of occurrence, building trend analysis data to allow operations to fine-tune processes, improve product quality, and increase production efficiency. Based on the Radiant TrueTest Software platform, these systems can simultaneously perform other inspections, including part orientation, assembly verification, fit and finish analysis, and label legibility, eliminating the need for a separate machine vision system on the line.

Photometry-Based Evaluation of “Just Noticeable Differences”

Because Radiant’s visual inspection systems are based on photometric software tools, they leverage advantages in light measurement for the detection and qualification of surface-level defects. One such advantage is the application of “just noticeable difference” (JND) for uniformity evaluation, which is the amount of change over a surface area in order for the difference to be noticeable at least half of the time. Radiant photometric imagers utilize JND to evaluate significant non-uniformities and mura (cloudiness) in illuminated displays, which would be deemed unacceptable by a human observer. As with illuminated components, JND measurement can be applied to non-illuminated surfaces for detecting and grading “uniformity” issues on surfaces caused by scratches, dents, smudges, and other defects.

Smudges detected on tablet screen using principles of mura or uniformity inspection.

A dent’s length and width can be measured to determine whether a device passes or fails.


Related Links:

Webinar: Visual Surface Inspection with Imaging Photometers