5 May 2021
With the advent of cutting-edge technology and sophisticated hardware, smartphone manufacturers constantly aim at the improvement of their products. In some cases, even small adjustments play a critical role in winning the completion among high-tech companies.
The main focus of a long-term Visidon’s client is on the unique features and performance of the camera. When they approached us to enhance the imaging quality, their Qualcomm-based smartphone platform was already strongly fine-tuned. However, they were still looking for better camera performance than their competitor’s flagship phone to achieve an outstanding imaging quality.
The primary issues were low-light imaging and noise reduction since these are common challenges for the end-user, i.e., product consumer. To address these challenges, we were involved in a multi-phase project with several development and evaluation stages between us and our client. To be specific, we first developed algorithms for offline testing and for simulation in PC environment.
After the preliminary evaluations and feedback, the algorithms were integrated into the mobile device for real-world testing. Then, we did further adjustments and fine-tuned the algorithm according to the client’s requirements. After a timely development and evaluation process, our client was satisfied with the final results. In an objective image quality assessment, the level of details, as well as noise reduction, were better than with the competing model´s camera. The phone camera also received great reviews highlighting amazing imaging quality in low-light condition.
Visidon has worked with this client for many years. Meanwhile, smartphone camera technology has evolved a lot. For instance, the number of cameras per phone has increased and the hardware is continuously improving, i.e., chipsets become more powerful handling more computational load, and other features, such as optical zoom and wide-angle lenses, have been introduced. Although the resolution and image sizes have also increased thanks to these advancements, the computation time for image enhancement must stay as short as possible. Consequently, the algorithms have to be fine-tuned to get the best out of the sophisticated cameras in a very short time.
Visidon’s denoising algorithm is based on multi-frame processing with optimized filters, as well as AI models (CNN-based noise reduction networks). The technology is optimized for Qualcomm and Mediatek mobile platforms and to ARM architecture. You can read more about the basic image enhancement techniques from our previous blog article.
Visidon denoising algorithm for low-light shot:
Visidon specializes in the development of fast and energy efficient image and video processing technologies. Our software products include for example automatic image quality enhancement, computational imaging, and face recognition.
Our technologies utilize the latest microprocessor architectures and machine learning algorithms especially for embedded industries. Our solutions can be found in over 1 billion mobile phones.