One of the main problems in photography is the blurred image occurring when camera is shaken. Restoring photo sharpness by the blurred image is the subject of fundamental investigation in the field of digital photography. In the course of research it was found out that defining the sharp image edges doesn’t necessarily result in correct estimation of the core. Under some circumstances it even makes it more difficult. To solve the problem there has been developed new metrics for measuring applicability of the data about image edges. A new effective method of estimating the core has been proposed. It is based on using the prior space distribution and improving the core by means of iterative support. Due to this it becomes possible not to use the threshold filter to the core elements. The results obtained show that restoration is high even when the image is heavily blurred. This differentiates the algorithm from the existing ones.