Ahead of his participation at Image Sensors Europe 2020, we spoke to Albert Theuwissen, Founder of Harvest Imaging about his presentation, his pre-conference masterclass, industry challenges and more.
Q. You will be speaking on Deep Trench Isolation during the main conference – can you give us some insight into what delegates will hear during your presentation?
Deep Trench Isolation of DTI is a feature that is part of the manufacturing process of modern CMOS image sensors. It has great characteristics : lowering optical cross talk, lowering electrical cross talk, avoiding blooming. In combination with other clever processing steps DTI can help in increasing near-IR sensitivity, and/or can even assist in creating pixels in the third dimension. The talk will give an overview of the wide range of advantages and applications of DTI in CMOS image sensors.
Q. You will also be running a pre-conference masterclass on ‘Developments In CMOS Image Sensors Since IS Europe 2019’ – why do you think that people should attend?
Because the masterclass will give a broad overview of new developments and inventions over the last 12 months. People who cannot follow all the new developments on a daily basis will get an overview in a 3-hour presentation. Not just new technologies and new architectures will be discussed, but also critical comments will be added to what is presented and published in the public domain, e.g. how realistic are the presented figures and data sheets ? A very similar, and successful masterclass was held at IS2019, so the concept will be repeated, but with completely new material.
Q. What do you think will be the biggest challenges to the image sensors industry in the next 12-24 months?
For the industry: how to make sure that the demand after image sensors can be met. There seems to be a shortage on production capacity these days. Even the largest manufacturing companies are seeking for production capacity outside their own company.
Q.What do you think will be some of the most interesting advancements image sensors in the near future?
The question whether new image sensor developments will be used for capturing higher quality images (digital photography, video) or for generating better data that can be used for further processing (autonomous driving cars, IoT, AR, VR, …). On the other hand, further improvements in sensor performance (irrespective of application) require more sophisticated technologies, resulting in very expensive investments.
Q. Who can still follow this trend?
Deep learning is one of those technologies that get attention everywhere, but recently also deep learning has shown some very promising benefits in image processing. E.g. very efficient demosaicing methods (which are independent of CFA architecture) are developed based on neural networks. This will give more freedom on the design of the CFA without having a negative impact on the image quality. In this case, the application of deep learning will open new possibilities to optimize the CFA for other purposes, e.g. to lower noise.
Q. What are you most looking forward to about attending the Image Sensors Europe Conference?
Meeting with people and peers of the imaging world. Having in-depth discussions about the technology that is being used all over the place : the image capturing devices.