To improve the reliability and sensitivity of an optical inspection system, the optical design of the sub-systems and components should be well-matched to both the inspection requirements and the optical properties of the product of interest. This is done by configuring the lighting, optics and geometry to match the relevant properties of the product to be inspected. Easy? The devil is in the details.

The basic steps to perform on-line optical product inspection and/or control are: first, point a light source at a part; second, sense the part with a photosensor(s) and/or imaging digital camera; and third, then send the detected light signal to a processor (digital or analog) to determine the pass/fail or other operations. Matching the optical properties to the design usually simplifies and speeds the signal and/or image processing as well as improving the reliability and sensitivity.

The most significant steps for matching the design to the requirements are to determine:

  • Function of the Inspection and Limiting Conditions: gaging, defect detection, structural integrity, etc. For example, the dimensional precision required determines the optical resolution necessary to achieve the performance.
  • Optical properties of the product part that best suit the function. For example, selecting the angular and spectral reflectivities of both the defect and the background surface can determine the possible sensitivity, dynamic range and Signal-to-Noise Ratio for defect detection and suggest the required geometry of the components as well.
  • If part consists of different materials: Inspection of sample for only one of the materials can be done when sample consists of different materials in different areas. For example, the inspection of products consisting of two different materials can often be simplified by selecting visible and/or non-visible (infrared and/or ultraviolet) light to match the optical properties of the material of interest (e.g. blue light for blue regions, and red light for red regions). Other differential properties can also be used (e.g. different directions of surface texture, geometry of part, etc.)

1. Limiting Conditions

Limiting conditions can reduce the design options and may require significant design changes. For example, the spectral properties of the light sources, filters and sensors should be able to match the spectral optical properties of the product part. The speed of the production line can limit the use of sensors which have slow response times.

2. The Product is a Critical Optical Component of the Sensing/Inspection System 

Exploiting the optical properties of the part to be inspected usually results in the best selection of illumination/lighting, sensors and geometry of the gaging system.

The detected light should contain useful information (i.e. Signal) about the product’s properties of interest. When possible, the detected light should not contain background or non-useful information (i.e. noise). Appropriate selection of the system’s components and the placement of components should improve Signal-to-Noise Ratio or S/N (and thus improve system performance).

Many optical properties, and applications, can be identified by visually assessing the product:

Is the inspected product transparent or opaque?

Transparent materials can be converted from and into opaque materials by selecting light wavelengths that take advantage of spectral absorption and transmission wavelength regions. If the product is opaque, the available volume can determine whether the detected light should be reflected from the product or be transmitted around the product.

 If opaque, is the light specularly reflected (i.e. mirror-like) or diffusely reflected, or provides non-uniform angular reflection patterns from surface texture patterns, or a combination of these?

This can provide the most suitable geometry for the incident light and for the detected light to improve the S/N by collecting the useful light (signal) that can be exploited such as differences in shape, texture axes, color, or periodicity of sub-components of interest in the product.

 If the product is transparent, is it clear, diffusing, have periodic internal or surface features, non-uniform?

All of the properties listed above can be, and have been, used to improve the performance of actual on-line systems.

These properties can be used for more complex applications.

Information specific to different materials in different areas of the product with different optical properties can use these different spectral regions (i.e. colors) to obtain selection of the different areas.

The geometry of the product and/or its components can be used for simplifying gaging and defect detection.

The different optical properties of different materials in different spectral regions are due to electronic and vibrational excitation that absorbs light for broad and limited spectral regions. Fundamental electronic absorption in solids results in light absorption at wavelengths shorter than a threshold wavelength, often in the ultraviolet. Vibrational absorption in solids often results in absorption bands in the infrared.

3. Advances in System Components: LED Light Sources

The most significant lighting advances have been by solid-state Light Emitting Diode (LED) light sources. LEDs have become available with both different colors and higher powers. Their use has been increasing dramatically.

Advantages and Development of LEDs

LEDs have the advantages of providing light without producing heat as with other sources. In other words, they are energy efficient, long-lasting (do not require similar replacement costs including parts and labor), rapidly switchable and are available to produce many colors including white.

For energy consumption: one label states that for the same eye-sensitive output of 800 lumens, 13 watts LED requires an input of only 13 watts compared to a 60 watts input for the same visible output in lumens from incandescent bulbs.

For long life, LEDs have a typical average lifetime of up to 50,000 hours compared to typical 1,000 hour lifetimes for 40 Watt incandescent lamps.

LEDs can be switched on and off in less than one microsecond, while incandescent bulbs take about ½ second.

The continuing development of the LED (which is a solid state device) is a result of increasing applications and sales volume in different commercial and consumer applications.

The increased use and associated production increases have resulted in decreasing unit costs, lower prices and better performance.

This relentless growth has been described by Haitz’s Law which states that the light output per LED has doubled about every two years since 1970 and the cost per lumen has been decreasing dramatically as well. This is similar to Moore’s Law describing the rapid development for integrated circuit electronics arising from expanded functionality and applications. This growth of both LEDs and integrated circuits has led to more development, applications and revenue.

LED growth continues.  Various estimates of the market size vary, depending on which products are included and the methodology used, but they generally agree that continuing growth will be about six to ten times as large in 2019 to 2021 as in 2013.  For example, Navigant Research forecasts that “annual worldwide revenue from LED lamps will grow from just over $1.5 billion in 2013 to more than $8.5 billion in 2021” while the revenues from LED lighting products, including both lamps and luminaries in commercial building markets will grow from $2.7 billion in 2013 to more than $25 billion in 2021.

In October 2013, the CEO of Philips announced that “LED-based sales grew 33% over the previous year” for the third quarter.

The third quarter total revenues for the company were said to have reached EUR 5.6 billion (US$7.56 billion).

 The LED growth has funded product development with features that can be useful for inspection and control.