Selective hydrogenation is an important process in petrochemistry to purify feedstock for polymer synthesis. For this process, catalysts containing metallic palladium deposited with an eggshell distribution on porous alumina are usually employed. For this kind of catalyst, the activity is known to be in close relation with the thickness of the palladium crust. As palladium oxide is brown and alumina is white, the palladium distribution in a catalyst bead before the reduction step can be characterized by optical microscopy. We propose an original and automatic procedure of optical image analysis to obtain a fast and robust method to measure the mean crust thickness of a catalyst batch and the corresponding standard deviation. The approach is validated by two different methods. First, we compared the crust thickness with those obtained by electron probe microanalysis. Then, catalytic tests of four samples with varying palladium crust thicknesses were performed and confirmed the expected correlation between activity and crust thickness measured by optical microscopy coupled with image analysis.