A system for detecting irregularities in three-dimensional parts being manufactured using an additive manufacturing process. Images of the manufactured cross section are obtained via an apparatus, then processed for irregularities as the object manufacturing cycle progresses. These images are processed through computational algorithms in order to identify areas of compromised integrity or quality. This data is then used to determine the risk of the part as manufactured, and an assessment is performed to determine if the process should continue. Should the manufacturing process be determined to proceed, the data is stored to be further assessed later by technicians, operators, and/or engineers.