In the semiconductor industry, the integrated circuit density is getting higher, which leads to a smaller margin of error for the bonding process. Therefore, accurate adjustments before bonding are often necessary. An accurate image alignment under a vision system is usually required to fine-tune the bonding position. However, a considerable depth of focus is sacrificed for sufficient resolution of the object, limiting the optics operation range. Digital holographic microscopy with digital refocusing is applied as an enhancement to extend the depth of focus. The proposed algorithm merges the image alignment method with a refocusing algorithm to extend the image alignment capability with digital holography. Tested with simulated images, the proposed algorithm showed no significant deviation to simulated parameters. Also, the alignment result is robust among several test patterns. Moreover, the overall time is reasonable for industrial application.
An alignment approach for data-bearing halftone images, which are a visually pleasant alternative to barcodes, is proposed in this paper. In this paper, we address the alignment problem of data-bearing halftone images on a 3D surface. Different types of surfaces have been tested , using our proposed approach, and high accuracy results have been achieved. Additionally, we also develop a data retrieval tool from an aligned image, in order to decode the data embedded in the original image. A system to assess the accuracy of alignment is introduced to quantify the effectiveness of the proposed alignment approach.