If a camera is calibrated for colour, the colours on a sRGB monitor largely corresponds to the colours as they appear in reality. With such as calibrated camera you are able to measure colour to standardised values which is particularly important in the printing industry. It means that colours in the image can be compared to standardised target values and this is essential for achieving a colour image true to the original.
Three Values of a Colour System – A colour system is a specific way of representing every colour in a colour space. Although there are many different systems , they all have at least three values which are required to characterise a specific colour. Some common colour systems include:
- sRGB – R(red), G(green), B(blue)
- YMC – Y(yellow), M(magenta), C(cyan)
- YUV – Y(luminance), C & V (chrominance)
- HSV – H(hue), S(saturation), V(value-brightness)
Four Steps of Colour Calibration – Using the Macbeth ColourChecker chart it is possible to colour calibrate a camera. This chart consists of 24 colour patches many of which represent as closely as possible natural objects, such as human skin, foliage and blue sky. The patches are made such that they reflect light the same way in all parts of the visible spectrum and so have consistent colour appearance under a variety of lighting conditions. This makes it perfect for industrial camera calibration.
Basler Ace cameras are a fine example of how colour calibration can be performed in machine vision cameras. They use the standardised four step approach; white balance, gamma correction , matrix correction and correction by a six-axis operator.
The end result is a camera that can be used confidently in colour inspection applications, print and visual inspection being very typical . It also means it is possible to discern and ‘separate’ very similar colours. This can be useful not only in print inspection but also for example pharmaceutical where drug/tablet colours can be very similar. With cameras that can be colour calibrated, it expands the use of industrial cameras into more automated inspection tasks where “true-to-life” applications have been very hard to solve without the use of human interpretation of colour.