The phenomenon of dynamic speckle is used for non-invasive whole-field detection and visualization of physical or biological activity in various objects through statistical description of laser speckle dynamics. Usage of 2D optical sensors to capture sequences of correlated 2D speckle patterns allows for building a pointwise estimate of a given statistical measure which should give a quantitative high contrast detailed 2D map of the spatial distribution of activity across the object surface. The aim of the present paper is to find out an effective way to enhance visualization of the activity map obtained by the normalized correlation-based algorithms. Similar to all processing algorithms, the built estimates exhibit strong fluctuations from point to point due to speckle nature of the acquired patterns. The fluctuations decrease the contrast of the built 2D activity map and worsen sensitivity and resolution of the dynamic speckle method. As a first task, we studied the distributions of the built estimates by processing of synthetic speckle patterns. As a second task, we applied smoothing to the activity map to achieve enhanced visualization. As a third task we considered building a map of a parameter related to the correlation radius of the temporal correlation function of the processes undergoing within the sample. The results are verified both by simulation and experiment.