In recent years, agricultural engineers working in research have been using modern modeling tools, such as artificial neural networks, with increasing frequency. This tool, as a universal approximator together with computer image analysis is used to create empirical models that describe phenomena and processes involved in extracting and processing plant materials. Artificial neural networks are able to generalize from acquired knowledge, and this is an important feature when analyzing data involving a large range of factors to determine a given process. The objective of this research work was to develop a neural model allowing the assessment of dried parsnip quality and its classification on the basis of digital photos. Obtained by the convection method, the dried parsnip was analysed and classified. Its characterisctic features were chosen, allowing classification according to quality. As the result of the research, a number of generated neural models were verified and validated.