Abstract: Rainfall deals with size of drops and amount of water precipitated from cloud in particular time. This method proposes the counting of the processing time, total amount of rainfall per second that is the rainfall rate, total number of rain drops, total volume of raindrops & direction of rainfall. It also represents the rainfall distribution in the graphical form along with the histograms details. In this methodology, the first stage is video acquisition stage in which videos of rainfall are captured by using two slow motion cameras having the same resolution for the identical amount of time. The required results are found out by using the non-sampled contourlet transform(NSCT), energy based image fusion ,a new iterative triclass thresholding technique in image segmentation that is based on Otsuís method. Human appearance would be neglected in this system due to which errors are minimized. The result which we are getting from this system will take very less time with the help of MATLAB. The proposed system is having greater accuracy than the existing methods in many aspects.

Keywords: Non-sampled contourlet transform, Otsuís method, Iterative triclass thresholding, Energy based image fusion..