Thesis Proposal - Image-Similarity Algorithms for Detection of Faulty Units in Production Tests

Background Electronic production-testing is the process of verifying proper functionality of manufactured electronics before they are shipped to customers or integrated into systems. Certain types of functionality require external cameras that capture an image of the device under test (DUT) and use that image to determine if the DUT is functional or not.

This can for example be achieved by matching images of a DUT in production against images of a working reference and thereby determining if the unit is working or not. Because it is very costly to approve faulty DUTs, the algorithms that process these images of produced units need to be accurate and robust to the varying conditions in a typical production-facility such as lighting.

Description of the thesis proposal The student will evaluate pros and cons associated with different image-similarity algorithms and evaluate them against an appropriate benchmark. The student might, for example, evaluate histogram-based solutions or CNN-based solutions. But the student is free to explore any class of methods they see fit or develop their own algorithms.

For successful thesis work, we believe that the applicants should have Basic knowledge of Signal Processing, especially Image Processing. Strong skills in Python for AI, data processing and mathematical computations. For example: Tensorflow or Pytorch, Numpy. Theoretical knowledge of Convolutional Neural Networks Basic knowledge of mathematical statistics Good knowledge in both Swedish and English, speaking and writing.

To give you the best possible support during your thesis, we’d like you to be able to come to the office connected to the project and spend most of your time working from there. Application: We look forward to receiving your resume, and preferably, a personal letter in which you explain why you want to write your thesis with Syntronic.

We screen and evaluate applications on an ongoing basis. The thesis project may be filled before the application deadline.