Count LED blink

edit

Hi,

The simplest way is to do a histogram matching (histogram matching opencv

For this, calculate the histogram for the ROI (blue rectangle) at this time, when the led is on (master template). Afterwards, using the master template to compare each frame within the ROI e.g., with Bhattacharyya, or Mahalanobis distance. If it is similar you have won!

However, there exists many solutions to do this and it depends on your requirements

(P.S.: when two classes have similar means but different standard deviations, the Mahalanobis distance would tend to zero, whereas the Bhattacharyya distance grows depending on the difference between the standard deviations Source: Wiki