OPPO is set to introduce OPPO Reno5 4G and Reno5 5G, the ultimate smartphone videography powerhouse, this February 2021. This marks the steady commitment of OPPO in keeping up with the changing times and needs of its young consumers especially in the advent of 5G connectivity. OPPO Reno5’s latest camera technologies will allow you to picture life together better with your family and loved ones.
OPPO has developed a more advanced and more comprehensive imaging system on the new Reno5 4G and Reno5 5G – the OPPO Full Dimension Fusion (FDF) Portrait Video System. Based on top-end imaging hardware, innovative image processing algorithms, and user-centric software, the OPPO FDF Portrait Video System caters to the diverse video creation needs of users in almost any situation, empowering them to capture the joy of life in every moment with easy-to-create professional- quality video content.
OPPO’s consumer research has shown that the core need for consumers is the ability to capture clear videos in any situation, and in a way that expresses their own unique style. At the same time, they want this process to be simple and easy to use.
The OPPO FDF Portrait Video System is comprised of two engines:
- Quality Enhancement Engine
- Portrait Perception Engine
Of the two, it is the Quality Enhancement Engine’s role to ensure users to capture clear images in any scenario, whether that be day, night, in strong backlight, while in motion or while standing still. The Portrait Perception Engine meanwhile, is responsible for establishing strict criteria for the features within a portrait video. These criteria ensure that the portrait subjects and the background are clearly and naturally defined.
The Quality Enhancement Engine is a suite of algorithms that ensure users can capture quality video images in any situation. Whether in low light, bright light, or fast motion environments, the engine optimizes settings to deliver images that stay true to life. It is the basis of many of the core imaging functionality on the Reno5 4G and Reno5 5G. Among these is OPPO’s original, industry-leading AI Highlight Video.
AI Highlight Video uses OPPO’s new generation, proprietary algorithm architecture to combine its industry-leading Ultra Night Video Algorithms and Live HDR Algorithms. By detecting light levels in a scene, it intelligently applies the algorithms to optimally adjust brightness, color and clarity in videos. This is the first time that OPPO has united Ultra Night Video and Live HDR under one feature, which is available on both the Reno5 4G and Reno5 5G front and rear cameras.
AI Highlight Video works by integrating three automatic optimization algorithms:
Adaptive Brightness Adjustment applies Auto Exposure (AE) - Adaptive Gain to control the exposure time and guarantee the brightness of the image while maintaining a smooth video from frame to frame.
Adaptive Dynamic Range applies dynamic range adjustments to each frame in a video with Dynamic Range Conversion and Filmic Tone Mapping technology to restore lost colors, brightness, and darkness to their natural levels.
Adaptive Noise Reduction introduces inter-frame fusion for temporal noise reduction and intra-frame fusion for spatial noise reduction.
To enhance the overall portrait capabilities of the device, the FDF Portrait Video System also includes the Portrait Perception Engine to enable users to shoot in their style. This ensures that no matter what the scenario the subjects are at, the Video Portrait System will process the portrait subject more naturally.
Scene Design Manager - a top-level portrait scene definition capabilities, together with OPPO’s self-built large database to analyze real-life portrait scenes and the processing requirements needed. It can define strict criteria for different scenarios and assess user-recorded videos against these criteria to provide suitable adjustments needed.
Portrait Recognition - After the Portrait Perception Engine has received the scene assessment from the Scene Design Manager, AI models and deep neural networks work to recognize the portrait subject in the image, separating them from the image background. To increase the accuracy and speed of the AI model, OPPO uses millions of portrait data from its own database to train the industry leading SOTA AI model, following which it conducts further processing to compress the model. With these optimizations, the whole portrait recognition process can be performed in less than a millisecond.
From beginning to end, the Reno5 4G and Reno5 5G leverage OPPO’s powerful technology to simplify the complicated task of video capture, editing, and sharing off of users’ hands, leaving them with a simple-yet-professional means to record and create their own stories.
.jpeg)
.png)
No comments:
Post a Comment