Theory of Light and Color
3. Visible Light
Broadly speaking, light refers to the ultraviolet, visible, and infrared range of the electromagnetic spectrum. This chapter explores the boundaries which define each range.
The Human Eye's Visible Spectrum
The sensitivity of the human eye to brightness is not always consistent across wavelengths. It peaks in the middle of the visible range around 555 nm and decreases as it moves further away in both directions until it reaches zero, at which point the eye will not feel brightness even if light enters. It is known that this sensitivity varies from person to person and even within the same individual over time.
The standard spectral luminous efficiency ≪1≫ is a standardized measure of the average spectral sensitivity of humans to brightness. In the above figure, the sensitivity (responsiveness) appears to be zero for wavelengths longer than 720 nm or shorter than 420 nm, but in reality, it is not completely zero. There is no clear boundary for the visible range, and its definition is not consistent even among academic societies and organizations. ≪2≫ One common range for human vision is from 380 nm to 760 nm. Despite the ambiguity, it is necessary to have these boundaries separating the visible range from the ultraviolet and infrared ranges.
Ultraviolet Light and Visible Light
The shorter the wavelength, the more harmful light is to living organisms because it holds more energy. The eye is especially sensitive, with the most sensitive part being the retina (visual cells) in the fundus, which is susceptible to damage from the strong photon energy of short wavelengths like ultraviolet. Tissues such as the cornea, lens, and vitreous protect the retina by absorbing these harmful wavelengths. This does not mean that UV radiation is safe to the eye though, as large amounts of UV can cause serious damage to the cornea and lens. The boundary between visible and ultraviolet defines the wavelength range that is safe for the eye.
Infrared Light and Visible Light
As described above, the human eye does not perceive brightness from light with wavelengths shorter than the visible range because of tissues such as cornea, lens, and vitreous. Wavelengths that enter the retina range from the visible to the infrared, at well over 1000 nm in wavelength. However, the human eye cannot normally perceive light in the infrared range at all. The main reason it is difficult to see infrared light is because the water that our eyes are made of absorbs infrared light at wavelengths just slightly longer than the deepest red color that we can perceive. Moreover, infrared light can’t activate the photoreceptors in the eye in most cases because it has less energy than the colors we see in the visible spectrum. Another reason for this phenomenon is illustrated in the graph showing the spectral reflectance characteristics of several plants.
The colors seen in the natural world result from the spectral reflectance characteristics of each wavelength in both the visible and infrared range. Spectral reflectance characteristics in the visible range differ for each color and therefore appear to fluctuate when expressed on a graph. However, the curve of the graph flattens out in the infrared range. This pattern in the spectral reflectance property appears for many different animals and plants.
Since ancient times, humans have survived on animals and the fruit of plants. Based on the color and shape of the fruit, they knew when it was ready to eat. As the fruit ripens, its color gradually changes. This means that the spectral reflectance property in the visible range changes even as it remains relatively unchanged in the infrared range.
The information in the natural world that humans must perceive to survive is concentrated mostly in the visible range. In contrast, the infrared range that humans cannot perceive does not greatly affect daily life. The boundary between the visible and infrared range is thus the result of human evolution.
Machine Vision's Visible Spectrum
The visible range of the human eye is approximately 380 nm to 780 nm. How does a machine’s eye, the camera, compare?
The most common cameras in machine vision are CCD and CMOS cameras. Most use an image sensor with a silicon-based semiconductor photosensitive element. ≪3≫
The photosensitive element converts light into electricity at a p-n junction. The sensitivity of the photoelectric conversion depends on the semiconductor materials.
Silicon-based photosensitive elements can detect wavelengths from approximately from 200 nm to 1100 nm, covering the human visible range. This makes them suitable for a wide variety of applications as the eye of the machine vision system. However, it is important to choose an imager according to the target wavelength range since silicon-based imagers have limited sensitivity to wavelengths outside of the human visible range.
In addition to silicon, the following types of detectors cover the ultraviolet range:
・gallium/phosphide (GaP)
・gallium/arsenide/phosphide (GaAsP)
To cover the infrared range, the typical detectors are as follows: ≪4≫
indium/gallium/arsenide (InGaAs)
indium antimonide (InSb)
As shown in the figure above, there are various types of photosensitive elements available from the ultraviolet to the infrared range to suit different machine vision needs. In many cases, you need to adjust the spectral response of the photosensitive element according to the machine vision application. For example, when evaluating brightness as viewed by the human eye, the total spectral response needs to be the same shape as the standard spectral luminous efficiency (λ) by applying an optical filter with appropriate spectral transmittance properties to the photosensitive element. ≪5≫
Another example is a UV germicidal lamp for bacteria. ≪6≫ Certain wavelengths kill or inactivate bacteria. A machine that evaluates bactericidal effects must have a spectral response that corresponds to the bactericidal wavelength.
When using a silicon-based element, it is important to note that it is sensitive to the visible and infrared range, as well as the ultraviolet range. Therefore, you need to adjust its spectral response to match the curve shown in the figure ≪7≫ by cutting the visible and infrared range with an optical filter. ≪8≫
Comment
≪1≫
Most academics have replaced the term standard relative luminous efficiency with standard spectral luminous efficiency.
≪2≫
| ISO | 360 ～ 830 nm （ TC 2 - 35 PHOTOMETRY draft ） |
|---|---|
| Law (Measurement Law) | 360 ~ 830nm (Measurement Law Chapter 19 Illuminometer Section 1 Verification Article 794) |
| Illuminating Engineering Institute of Japan | 360 nm or 400 nm to 760 nm or 830 nm (It is customary to use 380 nm to 780 nm for colorimetry) |
| JIS Z 8120 (Optical terms) | In general, the limit of the visible radiation wavelength range is 360-400 nm on the short end and 760-830 nm on the long end. |
≪3≫
The photosensitive element is a semiconductor electrical component that converts light into electrical signals. When light hits, the internal photoelectric effect causes the current and voltage to change in response to the strength of incident light.
Photodiodes, phototransistors, solar batteries, CCD image sensors, and CMOS image sensors are all common types of photosensitive elements.
In addition to semiconductors, phototubes such as photomultipliers utilize the external photoelectric effect.
≪4≫
This figure shows the wavelength bands in which various photoelectrical elements have their sensitivities, but there are differences in sensitivity even within a wavelength band.
≪5≫
Applying the optical filter on the photosensitive element is one method to optimize the spectral response property for a machine vision application. Another method uses spectroscopy, in which the light to be evaluated is divided into wavelength components (for example, 10 nm pitch), each of which is measured, multiplied by the function, and then added up. For an example, please see the calculation shown in ≪6≫ of Chapter 1.
≪6≫
As discussed in Chapter 2, the shorter wavelength, the greater the photon energy. UV-C, which has the shortest wavelength in the ultraviolet range, is harmful to living organisms. This consequently makes it useful as a germicidal lamp for killing bacterial species such as bacteria, viruses and molds.
Ultraviolet radiation near the wavelength of 260 nm as shown in the figure is widely recognized as an effective sterilizing method that works by destroying bacteria and the nucleic acids (DNA) of viruses. It is mainly used for medical purposes (sterilization of operating rooms, medical equipment), food (sterilization of ingredients and cooking equipment), cosmetic and pharmaceutical factories, etc. Furthermore, it is effective in preventing fish diseases by sterilizing breeding water, maintaining transparency of the breeding water (controlling algae), preventing odor, and spoilage of leftover food.
≪7≫
Source: JIS Z 8811-1968 Measuring methods of ultra-violet rays for sterilization
≪8≫
When using optics such as lenses and mirrors to design a vision system, carefully consider their various properties (spectral transmittance, spectral reflectance, etc.). The transmission and reflectance properties are important to consider especially when using ultraviolet or infrared light, otherwise it can often negatively affect the image.