This study investigates the dimensional accuracy and detectability of parts manufactured by Fused Deposition Modeling (FDM) technology through multispectral imaging analysis. Experimental samples fabricated from PLA (Polylactic Acid) and translucent PLA/PHA (Polylactic Acid/ Polyhydroxyalkanoates) were examined to assess surface quality, with a focus on the geometric precision of fundamental shapes. Specifically, the distance between opposite edges of squares was measured, and the diameter of circles was evaluated. A Keyence VS-L-160MX multispectral camera was employed using various lighting conditions, including UV, blue, infrared, and combined configurations, to optimize material transmittance analysis, contrast enhancement, and surface feature detection. High-contrast backgrounds were systematically chosen depending on the sample color to improve image sharpness and edge definition. Dimensional measurements were compared against nominal values, revealing that the greatest deviations approximately 10% occurred in the smallest geometrical features. These findings are attributed to the inherent limitations of optical resolution, calibration precision, and depth of field when analyzing fine structures. The study advances the methodology of surface quality assessment and underscores the necessity for optimized scanning parameters in the dimensional verification of additively manufactured components.