• DOWNLOAD

What accuracy can color sensors and mini-spectrometers achieve?

Spectrometers are used as reference devices in color measurements. However, they usually are to expensive and bulky for active usage in most applications. Color sensors and mini-spectrometers are effective solutions in this case. But which sensors and detectors are suited best for specific applications? Which factors, characteristics and accuracies need to be considered?

1. Introduction

Colorimetry or measurement of colors is becoming more and more important in modern applications. In the past simple RGB sensors have been used for some solutions. Nowadays the sensor or detector requirements are even more complex. Color measurement is a complex task and requires know-how and sophisticated sensor solutions depending on the respective application. The perception of color is not based on a physical values like current or pressure, but also includes a subjective or physiological element.

Each colorimetrical application requires an adequate system and sensor setup, including filter, electronics, lighting as well as a target for calibration to optimize the accuracy and output. The choice of light sources, the general system operation and filter characteristics decide what the sensor modules will be able to detect.

The electronics set the quality level of the sensor signal values and define the time frame and speed of measurement values. An ideal combination of light, sensor filters and amplifiers leads to ideal results. Missing or inaccurate calibration can lead to misinterpretation of the measurement values.

This paper intends to analyze the adequate usage of sensors and detectors (RGB sensors, True Color sensors, multi-spectral sensors and Mini-spectrometer) in relation to the application and demonstrate the finding via measurement values. The application-based choice of sensors will also be aligned with cost, speed, size and color accuracy characteristics.

2. Sensors & Detectors

2.1. Overview: Sensors & Detectors

Most commonly two different device kinds are used for colorimetric applications. First; the traditional spectrometer as reference and calibration device and secondly color sensors, as cost-efficient option with various resolutions and characteristics. Mini-Spectrometers are seen as an option. In the next chapters the various sensors and detectors will be used to perform test measurements in application-based setups to further evaluate the advantages and disadvantages of the types.

Image 1: Typical characteristics of RGB sensors based on absorption filters

RGB sensors

RGB sensors, based on absorption filters, consist of three band-pass filter in the visual range. The peaks of the spectral graphs are not set uniformly in relation to certain wavelengths but are defined during the manufacturing process based on measurement task and costs. One needs to know that this kind of color measurement is not based on human eye perception, however can or cannot be used in colorimetric tasks - depending on the required accuracy. Even with complex calibration methods, there are accuracy limitations.

 

Image 2: Typical characteristics of True Color sensors (XYZ) based on interference filter

True Color sensors

True Color sensors are used for absolute value color measurements. They use interference filter as a technological basis to implement color standards and are able to measure values more accurately than the human eye. The term "True Color" is in relation to the characteristics of a sensor type, which consist of filters featuring "better than human eye" standards, the DIN 5033 (or CIE 1931 XYZ color space).

These filters consist of a specific allocation of the sensitivity values of a color channel to the spectral wavelength. Deviations lead to a misinterpretation of the measurement values. Via calibration it is possible to determinate the color values as XYZ coordinates, which are used as base values for conversions into other color spaces. Based on the CIE 1931 standard observer characteristics, the True Color sensors measure absolute accuracy values, comparable to the ones of a human eye. Therefore it is possible to describe the color of fabrics or print products in the same way the eye would do.

 

Image 3: Typical characteristics of multi-spectral sensors, here with 6+1 characteristic lines

Multi-spectral sensors

If the spectral composition of objects needs to be measured or if metamerism effects are expected, the utilization of multi-spectral sensors is recommended. Multi-spectral sensors consist of a chosen spectrum, that is separated into spectral areas. The filters are arranged in such a way that their limiting ranges overlap, and almost no gaps in the visible spectrum exist.

For multi-spectral sensors the color evaluation takes place on the radio metric level and not the colorimetric level. As result one receives the spectrum of the sample and determines the color point via these values.

Mini-spectrometers

Mini-spectrometers are compact and sturdy solutions, which measure spectral values and allow an interpretation of the color space. The resolution is limited in comparison to common spectrometers. However based on this fact and the fewer spectral scanning spots they are faster than normal spectrometers but still slower, compared to RGB or True Color sensors

2.2.  Comparison

Color sensors are characterized by their small chip surface and high operation speed. They are capable of simple color comparisons up to absolute color measurements and radiometric measurements via multi-spectral sensors. Comparison means that one ore multiple samples are measured and used as reference values. A limiting value is set for RGB (relative measurements) or colorimetrical value YXZ like ΔEL*a*b* (absolute measurement) and used for further classification of measured samples. A requirement for comparison measurements is a known array of colors.

If the color point alone is not enough, one needs to use mini-spectrometer or multi-spectral sensors. These allow to perform measurements of an entire spectrum (up to a certain resolution) and compensate metamerism effects.

To compare the mentioned sensors and detectors various test setups based on actual applications (using the different sensor types and mini-spectrometers) will be used to demonstrate, when which kind of measurement characteristics are preferred.

3. Measurements

3.1. Application: LED regulation with color sensors

RGB and True Color sensors

More and more LEDs are replacing traditional lamps and are used as energy-efficient light source. Examples are display backlights, projectors or ambient lighting in for example in hotels, automation or aviation engineering as well as in medical applications. Often it is required to generate  mixed colors or specific color temperatures at a high Color Rendering Index (CRI). Additionally, effects like color shifts related to binning, temperature drifts or aging need to be compensated.

Color differences at a ∆ u’v’ ≤ 0.005 can be seen by the human eye. A trained eye can even visualize color differences up to a ∆ value of 0.003. The following test results describe the characteristics of RGB and True Color sensors based on the example of D65 and white light. Table 1 shows the D65 measurement values from the test setting.

 

Image 4: RGB sensor (absorption filter) with color drift. True Color sensor (interference filter) without.

Two systems using a feedback control loop with RGB and True Color sensors have been setup and calibrated at a temperature of 40°C (104°F). Afterwards the temperature of the LEDs has been changed which results in a color drift of the entire system. This drift was compensated by the feedback control loop. Table 1 shows that the RGB sensor has a control accuracy > 0.007 at 20°C (104°F) with further drift while increasing the temperature. However, the True Color sensor remains beneath visual range of the human eye - at a color distance of 0.0011.

Table 1: Comparison of RGB and True Color sensors during a D65 measurement

3.2.   Application: Display technologies

True Color sensors and Mini-spectrometers

In medical technology it is important that screens of diagnosis devices are capable of a high contrast ratio for detailed display options. Therefore display measurement devices with high precision and measurement sensitivity are required. To date display calibration was performed by calibration laboratories at high costs and efforts. With modern color sensors technologies and applications cost-efficient calibration system alternatives are available.

 

Image 5: True Color sensor in comparison with a mini-spectrometer (in Δ u'v')

The second test measurement series was performed by illuminating a diffusor plate with LED light at room temperature and 20°C (104°C) LED operating temperature and measuring the color point. With an application-based background the measurement values of a True Color sensor and a mini-spectrometer have been compared in reference to the values of a spectrometer.

The measurement results demonstrated that the sensors as well as the mini-spectrometer are capable of processing the signals faster than the reference spectrometer, however the error and accuracy values vary. The measurement values of the mini-spectrometer show a mean error of the color point measurement ranging from ∆ u‘v‘ 0.01 to 0.03 - hence are in the visual range of the human eye.

The measurements with the True Color sensor provides results with a mean error range of

∆u‘v‘ 0.001 - 0.005. Tolerance limits that are far below the visual range.

Table 2: Comparison of Mini-spectrometer and True Color sensors

3.3.  Application: Printing industry

True Color sensors and multi-spectral sensors

The printing industry has requirements for spectral value measurements during the printing process. In-line measurements can be a challenging task, since the measurement values are used directly to react and control the overall process.

For a practical test series a X-Rite ColorChecker was used for absolute color measurements. For this purpose a multi-spectral color sensor with an multi-channel transimpedance amplifier and flexible amplification levels was used. A white LED was used as standard light source. The multi-spectral sensor was used to measure the 24 spaces of the ColorChecker which have been compared with the reference values of a spectrometer. Via regression equation for spectral approximation it is possible to achieve average accuracies of ∆E00 = 0.72 (compare image 3).

Using a True Color sensor at identical conditions average accuracies of ∆E00 = 1.57 have been achieved.

An advantage of multi-spectral sensors is the greater accuracy and the possibility to use spectral approximation methods. If the printing colors are known, the results can be improved via calibration of the specific colors. Therefore it is possible to achieve absolute accuracies of ∆ E00 < 1 independent of the standard observer and standard light source.

The deviation values of the sensor compared to the spectrometer are ∆E00 = 0.3 for Cyan, ∆E00 = 0.9 for Magenta and  ∆E00 = 0.3 for Yellow.

Image 6: Evaluation of the measurement results via multi-spectral measurement

4. Summary

The evaluation demonstrated that True Color sensors are capable of achieving the accuracy of a mini-spectrometer regarding color measurements - in specific applications they even show superior accuracy. Therefore it is advised to review the specific application in the design phase to implement an adequate solution for the specific task at hand. It is necessary to know what kind of color or spectral information is required and how this data should be processed. The most suitable sensors or detectors are chosen depending on the required data, measurement procedure, accuracy and price.

For example: Mini-spectrometers are not capable of measuring consistent values for PWM controlled LED light measurements and therefore are inadequate for this kind of application. However, RGB and True Color sensors are not based on spectral measurements and therefore cannot be used in applications that require spectral values. In this case multi-spectral sensors or mini-spectrometer are the choice at hand.

Image 6 demonstrates a summary of the sensor and detector comparison. These results are based on a five-stage ranking: 1 Very high, 2. High, 3. Medium, 4. Low and 5. N/A.

The is a suitable sensor solution for every application. RGB sensors are a perfect match for simple color detection, True Color sensors are ideal for absolute color measurements and multi-spectral or mini spectrometer are suitable for absolute or spectral measurements.

Table 3: Overview of the analysis factors and interpretation

Authors

M.A. Kevin Jensen

Dipl.-Inf. Thomas Nimz

Bibliography

  1. MAZeT-Homepage: http://www.mazet.de
  2. Color sensor product information: http://www.mazet.de/en/products/jencolor
  3. Application note, Calibration of JENCOLOR® sensors based on the example of LED light sources, MAZeT GmbH, 2012
  4. F.Hailer, F. Krumbein, Application of JENCOLOR® multispectral sensors in dermatology, Ilmenau University of Technology (Thur.), 2011, urn:nbn:de:gbv:ilm1-2011imeko-083:6
  5. Walter Alt, Nichtlineare Optimierung, Vieweg+Teubner, 2002, 978-3528031930
  6. DIN5033. Farbmessung; Normvalenz-Systeme
  7. Friedhelm König. Die Charakterisierung von Farbsensoren. 2001
  8. Georg A. Klein. Farbenphysik für industrielle Anwendung. 2004.
Downloads
  • What accuracies can color sensors and mini-spectrometers achieve?
    View Download
Menu
top