We present a region-based approach for accurate pose estimation of small mechanical components. Our algorithm consists of two key phases: Multi-view object co-segmentation and Object pose estimation. In the first phase, we explain an automatic method to extract binary masks of a target object from multiple viewpoints. For initialization, the target object is bounded by the convex volume of interest defined by a few user inputs. The co-segmented target object shares the same geometric representation in space, and has distinctive color models from those of the backgrounds. In the second phase, we retrieve a 3D model instance with estimated upright orientation, and estimate a relative pose of an observed object in an image. Our energy function, combining region and boundary terms with the proposed measures, seeks the maximum in the overlapping regions and boundaries between the multiview co-segmentations and projected masks of the reference model. Our final results are high-quality cosegmentations consistent across all different viewpoints, accurate model indices, and pose parameters of the extracted object. We demonstrate the effectiveness of the proposed method using various examples.