Hi there,
Thank you for this excellent software. I am currently looking at the effect of the complexity of background in prey detection time in chickens. A complex background should make it difficult for the predator to detect the prey as more resources are needed to process the visual information in the background (Merilaita, 2003). Predation experiments with blue tits using artificial backgrounds have shown a higher number of contrasting elements, the complexity of element shape in the background (a more complex shape like a star compared to a circle), and the density of elements can all increase background complexity (Dimitrova et al. 2009; Dimitrova et al. 2010; Dimitrova et al. 2012; Dimitrova et al. 2014). I wish to use natural backgrounds for my study. But I am confused regarding what QCPA outputs to use, which could measure the complexity of the background. One way of measuring complexity is orientation clutter (Xiao and Cuthill, 2016). But I believe the visual acuity of the predator could play a role. So I created cone catch models of images of muddy background with few sandstones (perhaps a simple background) and areas with a lot more sandstones (perhaps a complex background) using the default sensitivities given for LMS peafowl (I do not have a full spectrum camera yet). I ran QCPA with the default weber fraction of LMS peafowl provided and used jungle fowls acuity value (I am unsure if this method is a good idea). I ran all pattern analyses. I am particularly interested in LEIA’s col/lum skew and kurtosis values. I read that higher values for skew could imply a more complex pattern, and low kurtosis could indicate the presence of salient features. I was wondering if these are reasonable measures of background complexity. Or if there is any other better measure of background complexity.
Thank you,

Measuring background complexity
Cedric van den Berg Answered question October 26, 2021