The coefficient autocorrelation is a descriptor used in many applications with biospeckle technique. However the analysis of results is through of interpretation of graphs curves. In this paper us proposal a fast algorithm for the linearizing of autocorrelation coefficient; through of the average of slope between autocorrelation coefficients and thus the computing machine can take decisions for many application; for example: fruit classification, maturity detection, storage detection, seeds classification etc.