This paper presents a visual inspection system for the orientation of micro-assembly components, designed for use in automated manufacturing complexes. The objective of the developed algorithm is to detect microchips placed in waffle packs that deviate from the permissible angular orientation. The proposed method involves image preprocessing based on Gaussian and multicriterial method smoothing, threshold segmentation, and morphological transformations, followed by object contour extraction. For each detected element, the rotation angle relative to the horizontal axis is estimated. Elements exceeding the set angular deviation threshold are registered and visually marked for subsequent positioning by actuators. The developed approach can be integrated into machine vision systems of industrial assembly lines to ensure orientation control and placement quality in micro-assembly processes.