<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>With the development of semiconductor assembly technology, the continuous requirement for the improvement of chip quality caused an increasing pressure on the assembly manufacturing process. The defects of chip pin had been mostly verified by manual inspection, which has low efficiency, high cost, and low reliability. In this paper, we propose a vision measurement method to detect the chip pin defects, such as the pin warping and collapse that heavily influence the quality of chip assembly. This task is performed by extracting the corner feature of the chip pins, computing the corresponding point pairs in the binocular sequence images, and reconstructing the target features of the chip. In the corner feature step, the corner detection of the pins using the gradient correlation matrices (GCM), and the feature point extraction of the chip package body surface using the crossing points of the fitting lines are introduced, respectively. After obtaining the corresponding point pairs, the feature points are utilized to reconstruct the three dimensional (3D) coordinate information in the binocular vision measurement system, and the key geometry dimension of the pins is computed, which reflects whether the quality of the chip pins is up to the standard. The proposed method is evaluated on the chip data, and the effectiveness is also verified by the comparison experiments.</p>
</abstract>

With the development of semiconductor assembly technology, the continuous requirement for the improvement of chip quality caused an increasing pressure on the assembly manufacturing process. The defects of chip pin had been mostly verified by manual inspection, which has low efficiency, high cost, and low reliability. In this paper, we propose a vision measurement method to detect the chip pin defects, such as the pin warping and collapse that heavily influence the quality of chip assembly. This task is performed by extracting the corner feature of the chip pins, computing the corresponding point pairs in the binocular sequence images, and reconstructing the target features of the chip. In the corner feature step, the corner detection of the pins using the gradient correlation matrices (GCM), and the feature point extraction of the chip package body surface using the crossing points of the fitting lines are introduced, respectively. After obtaining the corresponding point pairs, the feature points are utilized to reconstruct the three dimensional (3D) coordinate information in the binocular vision measurement system, and the key geometry dimension of the pins is computed, which reflects whether the quality of the chip pins is up to the standard. The proposed method is evaluated on the chip data, and the effectiveness is also verified by the comparison experiments.

Automatic Detection of Chip Pin Defect in Semiconductor Assembly Using Vision Measurement

Measurement Science Review

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

With the development of semiconductor assembly technology, the continuous requirement for the improvement of chip quality caused an increasing pressure on the...

Automatic Detection of Chip Pin Defect in Semiconductor Assembly Using Vision Measurement

Online veröffentlicht: 05. Aug. 2022

Seitenbereich: 231 - 240

Eingereicht: 25. Nov. 2021

Akzeptiert: 22. Mai 2022

DOI: https://doi.org/10.2478/msr-2022-0029

Schlüsselwörter
3D reconstruction, chip pin, defect detection, feature extraction, computer vision

© 2022 Shengfang Lu et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Automatic Detection of Chip Pin Defect in Semiconductor Assembly Using Vision Measurement

Online veröffentlicht: 05. Aug. 2022

Seitenbereich: 231 - 240

Eingereicht: 25. Nov. 2021

Akzeptiert: 22. Mai 2022

DOI: https://doi.org/10.2478/msr-2022-0029

Schlüsselwörter3D reconstruction, chip pin, defect detection, feature extraction, computer vision

© 2022 Shengfang Lu et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Schlüsselwörter
3D reconstruction, chip pin, defect detection, feature extraction, computer vision