Migration successful. You may run into the following message when using git:
WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED!
This message is legitimate and you can safely update your known_host file.
Many methods can be used to construct the force chain network from polarized images. For low resolution imaging, where individual contacts are not detectable, one can use a specific filter called ridge detection. Ridge filter uses gradient of a smoothed version of the image to find the maxima of intensities. These maxima correspond to force chains. Using the gradient, one can also find the direction of the force chains. Ridge filter can be implemented in [Matlab](https://www.mathworks.com/matlabcentral/fileexchange/63171-jerman-enhancement-filter), [Python](https://stackoverflow.com/questions/48727914/how-to-use-ridge-detection-filter-in-opencv), or using builtin function in [Mathematica](https://reference.wolfram.com/language/ref/RidgeFilter.html). Figure below shows a sample image with ridge detection on it. The color represents intensity from the image from blue being black to red being white. With this filter, all background noise from particle edges can be removed and the network structure can be constructed more easily.
* Showing force chain using ridge filter. a) A raw polarized image from the experiment. b) Processed image using ridge filter and color coded with image intensity. Blue to red corresponds to black to white respectively, showing the force chain strength: