Digital imaging is a powerful tool for measuring the visual world around us, but it comes with a range of pitfalls and limitations. The micaToolbox and QCPA framework offer a software suite for objectively measuring colours and patterns, and being able to do so in animal colour-vision.
Taking Photos/Digital Images
Image acquisition is the first, and critically important step. There are many factors to consider when taking images, particularly the equipment and lighting (e.g. direction and type of light source), and control of any aspects of the imaging protocol which might influence the thing you’re measuring. If you need to work with animal vision sensitive to ultraviolet light then there are additional considerations, likewise for underwater photography. As a rule it is best to use RAW images from digital cameras, but the toolbox does support non-linear images such as JPGs.
The images now need to be calibrated to control for fluctuations in light intensity and colour, and (if using JPG or other non-linear images), linearisation, to ensure the pixel values scale linearly with radiance (note cone-mapping is also recommend when using non-linear images, and is sometimes essential depending on colour-space). Following calibration the image pixel values are scaled to percentage reflectance (in red, green and blue, in camera colour-vision) relative to the grey standard(s) in your image.
Mapping to Animal Colour-Vision
Each camera/lens combination has its own “colour-vision” (spectral sensitivities), which will not match those of human colour-vision or any other animal. Many colouration questions in biology should be addressed using the colour-vision of the receiver. The toolbox therefore offers “cone-mapping” functions to create digital images where pixel values describe the relative cone-catch quanta of each receptor.
The toolbox is bundled with a large range of image analysis and visualisation tools. The QCPA framework offers a streamlined approach where you can control for receiver acuity and viewing distance, and then extract a large range of descriptive statistics from every image.
Additional analysis tools include measurement of edge disruption (GabRat) and pattern energy analysis (also called granularity and bandpass analysis).
Further tools allow you to visualise colour information in various ways, such as colour maps, presentation images, and RNL chromaticity/saturation images the latter is also a useful method for measuring “hue” and “saturation”.
Please look through the user guide for up-to-date information on all the features, and see our peer-reviewed papers for additional overviews, details and citations. If you have any outstanding questions please watch the videos(!) and use the the forum to ask anything else (we’ll reply faster than via email, and others can benefit from the answers).