Sketches are the main tools for the communication of concepts among design team's members during the ideation phase of the design process. Imprecisely defined sketches lead to uncertainty in communication during the design process. Thus, as a contribution to reduce the uncertainty in design communication, an initial framework for the quantification of uncertainty associated to sketches was presented in previous work. In that initial framework, the probabilities of the features in a sketch were determined based on the assessment of an experienced designer. This approach reduced the usability of the framework by professionals with limited experience e.g. entrylevel engineers. Thus, this posed the need of an improved framework and brought the following research question: Can a probabilistic method be used to improve the quantification of uncertainty in sketches? Accordingly, to answer this research question the following specific aims were established: 1) Ranking of features in a sketch, 2) Determination of the probability of importance of features in a sketch, and 3) Quantification of uncertainty in a sketch. The first aim focused on determining and classifying the features in a sketch, based on a hierarchical approach. The second aim focused on determining the probability of importance of the features in a sketch, by assessing its probability of likeliness using an object recognition approach, and by applying a probability transformation. The third aim focused on the quantification of the uncertainty in a sketch, based on the calculation and normalization of the sketch's entropy. This resulted in the development of an improved framework for the quantification of uncertainty in sketches, which can be used by design practitioners with limited experience, and whose application is presented and detailed in a case study.