Glassless 3D technology using
AI-technology from 2D to 3D image conversion in real-time
Glass-Free 3D technology by AI real-time 2D to 3D image conversion
（Currently, Patent Pending)
Related technology: NOVIUS proprietary technology: Patented (AI image platform, real-time style transfer)
NOVIUS successfully developed AI technology and solutions deliver stunning 3D stereo images in real-time from monaural image taken by monocular camera that creates to stereoscopic images. This leads to see high quality 3D video image by glass-free display in real-time. The name of this AI solution is called “N-VISION”
NOVIUS developed proprietary AI achieved real-time output of stereo images (side-by-side) from monaural images that are extremely close to human stereo-
scopic vision that leads to glass-free 3D image and video.
In the medical field, it will facilitate diagnosis and understanding of lesions that require stereoscopic interpretation, and contribute to improved surgical skills and see 3D images based on depth and realistic distance can be shared by inexperienced and residence doctors simultaneously with glass-free monitor in the operating room.
We have achieved world’s fastest conversion speeds with NVIDIA- installed PC available on the market. this is the Victory of our AI technology.As for the time
Creation of depth map using AI -> 0.01 sec. (Auto Optimization Function)
Creation of L→R→0.008 sec.
Drawing Side-by-Side to the screen → 0.007 sec.
What does it means to be real-time conversion from 2D to 3D
Requiring real-time stereoscopic image and video needed in various fields.
What are the advantages of converting from 2D (flat) to 3D (stereoscopic) in real time, and how does the conversion effect to our life? Does it mean a progress of technology? What is the difference between what was seen in 2D and what would be seen in 3D view? What I can certainly say is that we can be seen things that had not been visible as 2D by making them 3D in real time. Understanding the situation in real time 3D conversion can even change the outcome. For example, in the medical field, during operation, viewing depth that is possibly seen lesion site for tumor and polyp that are hiding can be found. Real-time 2D-3D in infrastructure inspection cameras can be applied to aircraft jet engines, where very small cracks on the millimeter scale can lead to fatal accident
Regarding 2D-3D Conversion Technology
The recent development of 2D to 3D conversion technology is being developed in USA and Chia and making extensive use of artificial intelligence. China and the U.S. are particularly active in this area, and the technology has been applied to medical and animated and video images. what about real-time conversion of these existing technologies? At present, NOVIUS is the only company in the world to offer Glass-Free 3D technology that use side-by-side image in real-time 2D to 3D conversion.
Current speed of the conversion is about 50 nano seconds from 2D to 3D as a whole.
Patent Pending (Own 100% of the intellectual property rights)
N・VISION 3D technology incorporates AI with the proprietary algorithm developed by NOVIUS alone. Accuracy of the model (training data) can be increased and the depth will be deepened. AI is generated from a normal 2D image to predict the distance within the area where depth is in essence. This distance is called DEPTH. Although not included in the patent application, AI also corrects and adjust the side-by-side position automatically.
N-VISION3D Technology's Strengths
Differences between similar technologies
Similar technologies can be found today in terms of real-time stereoscopic representation of real-time video and image.
A technology that uses lasers to project stereoscopic images into space; no need 3D glasses; viewable with the glass-free eyes; holograms record the three-dimensional image of an object as it is, so they can be seen not only from the front but also from the side or back, which is very different from other imaging technologies. The technology used to create holograms is called holography. The original essence of this technology is to take advantage of its characteristics as an image recording rather than an image display. Of course, the principles used are different, but the major difference between the above technology and our technology this time is the ability to convert real-time or not.
Glass-Free 3D Display
About 10 years ago, major TV manufacturers launched 3D displays under the name of “3D display. In the case of stereoscopic TVs, the contents to be viewed were limited. This is because all visual contents that could be viewed with glass-free 3D display. Special glasses had to be created by TV broadcasters and production companies. In addition, there was a huge production cost involved in 3D live broadcasts, including special equipments, and other major factors related to video contents. The major difference between the above technology and NOVIUS AI technology is that it can be shared by multiple audiences without the need for glasses. Again, real-time and immediate 3D visualization of live-action images is possible.
The major difference between these three technologies and NOVIUS technology is that it is difficult to share with multiple people. In particular, with VR technology, the user wears goggles, so only the person wearing the goggles can see the image information at that time. For MR technology (an advanced version of VR and AR), there is a technology called “Microsoft HoloLens” which is a 3D display of a specific animal using a specific application
Examples of Technology and Application.
Semiconductor (Back Process)
To carry-out high-performance computing, broadband 5G network switching, autonomous driving, integrated sensing, AI diagnostics, and so on. Further integration and performance enhancement of semiconductor devices is required.3D packaging technology (technology to densely mount logic, memory, and peripheral devices in a single package), for which real-time 3D image check-up is in essence required
Using N-VISION, we aim to create 3D images in real time from monocular camera, and create images where infrastructure inspections can be performed only by drones and robots. This technology can precisely measure the size of cracks, including length, area, and volume on exterior walls, etc., using only the information obtained from the monocular camera image. Camera equipped with DRONE or ROBOT does not need to change it.
3D industrial endoscopes providing a 3-dimensional picture, which supposedly gives additional information of depth, mechanical details and orientation in the field.
Actualization the world's fastest real-time 2D-3D conversion time
We have developed our own AI 3D algorithm to achieve the world’s fastest 2D-3D real-time conversion speed. There are several factors that made this possible.
As a specific conversion speed, the total time required to upload a 3D image is around 0.05 seconds (With 24inch Lenticular display)