Deformation is a phenomenon that is common during drying. Generally, deformation is evaluated and reported in terms of volumetric shrinkage. However, volumetric shrinkage cannot be used to describe non-uniform deformation, which commonly takes place during drying of shrinkable materials such as biopolymers, fruits and vegetables. In the present study, algorithms and software were developed to first reconstruct 3D images from 2D images, and then to monitor non-uniform deformation of such images. Agar gel with different sugar contents were used as the test materials and were allowed to undergo drying at different rates. The developed image reconstruction algorithms were successfully validated with standard geometrical shapes. Various image-based parameters, both in two dimensions, i.e. projected area, perimeter, major axis length, minor axis length, equivalent diameter, extent and fractal dimension; and in three dimensions, i.e. volume, sphericity, Wadell's sphericity, Wadell's roundness, radius ratio and Hoffmann shape entropy, were then calculated. The results showed that extent and fractal dimension could be used to identify the onset and to monitor non-uniform deformation in two dimensions, while sphericity could be used for such tasks in three dimensions.