What is colour accuracy? How to measure colour accuracy?

Accurate representation of colour is essential for identifying and distinguishing objects or people, product inspection, colour matching and for seamless quality-control. Colour accuracy is often assessed using standardised colour charts or targets. It is measured in terms of Delta E (ΔE), which quantifies the difference between the displayed colours and the reference colours. A lower Delta E value indicates better colour accuracy.

Colour accuracy can only be achieved by maintaining white balance and saturation at optimum levels. The colour gamut, sensor characteristics, lens quality, calibration etc. are also other factors that influence it.

In this blog, we are exploring the factors that determine colour accuracy and measures increase the same.

What is white balance? How does it influence colour accuracy?

White balance is a camera setting that establishes the true colour of white. This produces a baseline from which all other colours are measured. White may not appear “white” under all lighting conditions, so setting the white balance right helps correct this anomaly.

Achieving white balance (WB) involves removing the unrealistic colour casts so that objects that appear white in person are rendered white in the photo. Proper camera white balance has to take into account the colour temperature of a light source, which, in simple terms, refers to the relative warmth or coolness of white light. Let’s understand colour temperature better.

Understanding colour temperature

Colour temperature refers to the characteristic colour of light, often measured in Kelvins (K). It’s a way to describe whether a light source appears warm or cool. Where the warm light and cool light are differentiated as follows:

• Warm Light (Lower Kelvins): Light with a lower colour temperature, such as the warm hues of a sunrise or the light from incandescent bulbs, appears more reddish or yellowish.
• Cool Light (Higher Kelvins): Light with a higher colour temperature, like the light on a cloudy day or from fluorescent bulbs, tends to have a bluish or cooler tone.

Understanding colour temperature is crucial for achieving accurate and appealing colour reproduction in imaging devices. The white balance setting on the camera can be set to account for different colour temperatures in various lighting conditions. For example, setting the white balance to a lower Kelvin value (warmer) when shooting indoors with incandescent lighting can help neutralise the orange tint often associated with such lighting.

Human eyes are very good at judging what is white under different light sources, but embedded vision cameras often have great difficulty with auto white balance (AWB). This can create unsightly blue, orange, or even green colour casts.

What is saturation? How is it correlated to colour accuracy?

Saturation refers to the intensity or vividness of colours in an image. Highly saturated colours appear vibrant, while desaturated colours appear more muted or grayscale. Lowering the saturation of a photo can have a muting or calming effect, while increasing it can increase the feel of the vividness of the scene. It is important not to over-saturate a photo, as sometimes it creates unnatural colour spill-over effect as shown in the following images.

In applications like surveillance or medical imaging, where accurate representation is crucial to the decision making, maintaining an appropriate level of saturation helps in conveying visual information accurately without introducing unnecessary colour distortions.

Saturation, when properly controlled, contributes to colour accuracy. It represents the richness and vibrancy of colours as they appear in the real world. However, excessive saturation can lead to artificial and unrealistic colour representations, impacting overall accuracy.

Remember, while accurate colour reproduction is the primary goal, it’s also important to consider the intended aesthetic or purpose of the imaging.

Given below are other additional factors that impacts colour accuracy:

• Colour Gamut: The colour gamut of an imaging system refers to the range of colours it can reproduce. A wider colour gamut allows a more accurate representation of a broader spectrum of colours.
• Colour Space and Profiles: Colour spaces define the range of colours that can be represented, and colour profiles provide instructions on how to interpret and display colours within a given colour space. Understanding and adhering to industry-standard colour spaces and profiles ensure consistency and compatibility across different devices and applications.
• Sensor Characteristics: In the case of cameras, the characteristics of the image sensor play a crucial role in colour accuracy. The sensor’s spectral sensitivity and noise levels can impact how well it captures and reproduces colours in different lighting conditions.
• Lens Quality: The quality of the camera lens affects the transmission of light and can introduce optical aberrations. High-quality lenses help maintain colour accuracy by minimising distortions and ensuring that light is properly focused onto the image sensor.
• Lighting Conditions: The lighting conditions during image capture or display significantly influence colour accuracy. Different light sources have distinct colour temperatures, and variations in lighting can introduce colour casts.
• Calibration and Profiling: Regular calibration of imaging devices is necessary to ensure consistent colour accuracy over time. In calibration, the colours on our device’s monitor is measured and tuned with a standard set of colour chart.
• Image Processing Algorithms: The algorithms used for image processing, including colour correction and enhancement, can impact colour accuracy. Careful consideration of these algorithms and their parameters is necessary to avoid introducing artifacts or distortions.
• Display Technology: For display devices, the technology used (e.g., LCD, OLED, LED) can affect colour accuracy. Each technology has its characteristics, such as contrast ratios, and colour uniformity, which can influence how colours are displayed.

How to measure colour accuracy?

Given below are the parameters that depict colour error:

• ΔE (Total Colour Difference): It’s a measurement of how much the captured colour varies from the reference colour. A lower Delta E means better colour accuracy. ΔE is calculated using formulas such as CIEDE2000 (CIE Delta E 2000) or CIELAB (CIE 1976 Lab*). CIEDE2000 and CIE 1976 (Refer Figure 4) are commonly used formulas for calculating colour difference that takes into account differences in lightness, chroma, and hue. A ΔE value of 0 would mean a perfect colour match.
• ΔC (Chromaticity Difference): It’s a measurement of how much a captured colour can differ from its reference colour by eliminating the luminance. ΔC is calculated by eliminating the luminance component and considering only the chromatic differences between the captured colours and reference colours. ΔC is usually calculated using the difference in chromaticity coordinates such as Δa*, Δb*, Δu’v’, etc.

e-con Systems’ cameras that help achieve colour accuracy

e-con Systems has 20+ years of experience in designing, developing, and manufacturing OEM cameras. One of our USPs is the ability to customise camera solutions to fit the exact needs of the clients, no matter the type of embedded vision system they are looking to build.

We also understand that image processing plays a vital role in ensuring the end image quality. Our team of experts ensures that every step can be taken for a fine-tuned processing, including gamma correction.

This article was originally written by Prabu Kramar, CTO and Head of Camera Products, e-con systems.