A new image analysis method for the spatial characterization of microscopy images of fruit microstructure is proposed in order to analyse the heterogeneous microstructure of unprocessed fruit and the possible inhomogeneous effects of various technological treatments on this microstructure. The microstructure of tissue samples was characterized using the global statistics of size and shape parameters calculated for all visible objects. Global analysis was supported by a novel algorithm that allowed for drawing of the maps of the cell wall fraction from microscopy images and for the analysis of both global and local compaction or loosening of tissue. The spatial distribution of the cell wall fraction was visualised in the convenient form of bivariate histograms. To test the developed image analysis protocols, structural changes resulting from ultrasonic and osmotic treatments of apple tissue samples were studied. Peeled and cored apples were submersed in a liquid medium (distilled water or 60 oBx sucrose solution) for 45 and 90 min with and without ultrasonic treatment. After these treatment procedures, tissue samples were cut into slices, stained and imaged using a microscope. The proposed method allowed to characterise the effects of different sample treatments.