The objective of this paper is to develop a model of an image compression, to make simulation and analysis by using Upgraded Petri Net (UPN). This paper describes an Upgraded Petri Net model which can generate results as compressed image for given input data of image and initial marking. The UPN presents formal modeling tool, based on mathematical apparatus, used for simulation and analysis of processes, particularly at the register transfer level. Original software for modeling and simulations of UPN, PeM (Petri Net Manager), is developed and used for all models described in this paper. This paper includes a preface on the UPN theory, the UPN models formulation, their simulation and analysis. The first model is a start model of an image compression. From this model we can derive a new model which includes elements we want to simulate and analyse. The second model can generate data we want to analyse in order to make new initial marking of the model or to change the model. The model implements a sub-image JPEG compression by using sequence of split image into blocks, discrete cosine transformation, quantization and entropy encoding. By setting the initial state of the model, we can simulate and analyze the different parameters. The results are stored into the *.mem file inside the PeM. The suitability of UPN for modelling of the JPEG image compression is examined and established. All models were developed by using original program suite PeM (Petri Net manager).