SingaporeMedia Processing Group
PSL Media Processing Group has been working on next-generation futuristic research and development on image/video technology development. Our group has been contributing very much to the image/video processing in Panasonic.
Our R&D activities
PSL camera team has been built-up since 2006 as we believe our Expertise in image processing and recognition can be applied to our camera business and moreover we have the Passion to develop World First/World Best features for next generation camera.
The Camera Technology Team Motto
“Towards Global Excellence -
Discover the nature & Apply to future camera product”
Our Technology
We have developed several technologies:
· Motion Blur and Ghost Removal
· Flickering Reduction
· Bayer Raw Noise Reduction
and now we are working in the area of Image Alignment, all-in-focus/Refocus and High Dynamic Range for future camera.
Motion Blur Removal [Patent Pending Technology]
Conventional
|
With PSL Motion Blur Removal (*Actual results shown here) |
Blur caused by moving objects is a well-known problem. We have developed a robust and low complexity method to solve the problem. Our unique point is pixel adaptive motion detection utilizing sensor noise characteristics; which makes our algorithm invariant to noise condition in the scene. Collaborating with Panasonic Platform Development Center (PFC) and PSL LSI Team; we successfully ported the technology onto FPGA (@50MHz) with 150k gates, to achieve more than 100fps.
Flicker Reduction [Patent Pending Technology]
(*Actual results shown here) |
Flicker is caused by Fluorescent lamp which uses Alternating Current. There are various existing methods to solve this issue. However, we find none which is suitable for next generation high speed camera. Hence, we developed our own technology to over come this.
Noise Reduction
Original
|
With PSL Noise Reduction (*Actual results shown here) |
In many camera systems, noise is the major problem in low light condition. We target to develop state-of-the-art noise reduction algorithm for ultra-high ISO capturing.
|
|
Our R&D mission is to create top performance image recognition technology for PANASONIC future and current products. Especially, we target on technologies realizing applications related to the keywords “Ease of Use” and “Safe and Secure”.
We focus on image-related statistical learning, pattern recognition and computer vision. Our key technologies are:
1. Face detection and related technologies like face recognition, age/expression estimation and face features (eyes, nose, mouth) analysis
2. Human detection, behavior recognition and hand gesture recognition.
3. Object/Scene recognition
Our face detection has been deployed in PANSONIC LUMIX digital cameras. We are now deploying image recognition technologies into next generation of ubiquitous sensor network devices like intelligent surveillance system and car safety system, including pedestrian detection and driver monitoring. We are also looking into new applications of media content management like photo-tagging or video search and retrieval system. In future, we plan to work in areas of new human-machine interface like interactive digital signage and applications of pattern recognition for intelligent image processing.
|
|
Pedestrian Detection for front sensing in-vehicle camera and Driver Monitoring for inner sensing in-vehicle camera |
|
Since 1992, PSL image/video team has been working on image/video technology development involving international standards such as MPEG-1, MPEG-2, MPEG-4, MPEG-4 AVC/H.264, MPEG-4 SVC, MPEG-4 MVC, and China AVS. Over the years, our team has built a reputable position as one of the main contributors in image/video processing in Panasonic.
Our R&D activities
| In recent years, we involved in MPEG-4 AVC/H.264 video coding standardization, which won Emmy Award. This standard is now used widely in the world from low-bitrate Internet streaming applications, high-definition HDTV broadcast and Blu-ray Disc applications, up to Digital Cinema applications with near-lossless quality. At the same time PSL utilizes the accumulated technology know-how to contribute to Panasonic AV product developments |
Before |
|
 |
|
After |
Present R&D activities of the PSL image/video team include making technical contributions to the development of next generation international standards and video coding algorithms for future applications such as 3D video and high-presence video. Through our efforts, we strive to fulfill our mission of providing realistic and immersive viewing experience that brings the world to every home. |
“Healthcare” is one of the biggest matters of concern for both individual and society and its importance is expected to increase in the future. There are many aspects in healthcare, such as prevention, screening, treatment, and health promotion, but we think bio-sensing is a common core technology since it is always necessary to know condition of each individual. Especially, we focus on optical imaging and spectroscopy as a non-invasive sensing modality and try to develop new sensing systems by combining optics with our signal processing technology accumulated in AV product’s R&D. Now we are building an optical tomography system in our lab collaborating with a university partner and also exploring new sensing applications which utilize optical characteristic from the interaction with human tissue.
