Systems and methods facilitate iteratively inspecting a high resolution image of a camera subassembly glass element, to determine whether dust is present, and cleaning of same. By examining the size, number, and location of the dust particles it can be determined whether the dust is likely to cause issues when the camera subassembly is integrated into a fully assembled document scanner. Other anomalies, contaminants, and defects which could affect camera image quality, such as scratches on the glass covering, can also be detected. Software may control the capturing of the high resolution images of the camera subassembly glass and can determine a fail state if the anomaly is above a predetermined threshold. The systems and method can be used as acceptance criteria of individual camera subassemblies from a manufacturer.