The distribution of pixel values in an image can give us valuable information about the image, for example if it is a positively skewed distribution, we can see that there is significant data in the image. If the distribution is roughly symmetric, we can tell that there is no significant data in the image.
On the other hand, in some measurements that we do on the image, we might need to know the certain statistical parameters of the image. For example, if we have run a detection algorithm on an image, and we want to see how accurate it was, one method is to calculate the average of the undetected pixels and see how reasonable it is (if detection is done correctly, the average of undetected pixels should be approximately equal to the background value, see Sky value). ImageStatistics is built for precisely such situations.
|• Histogram and Cumulative Frequency Plot:||Basic definitions.|
|• Sigma clipping:||Definition of \(\sigma\)-clipping|
|• Mirror distribution:||Used for finding the mode.|
|• Invoking astimgstat:||Arguments and options to ImageStatistics.|