?빠르고 지능적인 동영상 반도체 IP 

 

1. 인공지능 기반의 영상의 품질을 향상시켜주는 영상 화질 개선 및 업스케일링

자체 개발한 AI가 저해상도 영상의 화질을 개선하면서 해상도를 높여줍니다. 수많은 연산이 필요한 동영상 처리 AI를 반도체 기술로 구현해 빠른 속도로 영상을 처리 할 수 있으며, 실시간 화질 개선 및 업스케일링이 가능합니다(SD to FHD).

 

2. 인지 화질은 유지하면서 영상의 용량을 30% 줄이는 인코딩 전처리 AI

동영상 인공지능 모델이 화질에 영향을 주지 않는 화소를 구분해 용량을 30% 줄일 수 있습니다. 인코딩 전처리 단계에서 영상을 처리해 어떠한 코덱에도 적용 가능하며, FPGA 카드 한 장으로 4Kp60 실시간 처리가 가능합니다.

 

3. 고성능 AV1, AVIF 코덱

기존 h.264, jpeg보다 압축 효율이 높아 차세대 코덱으로 기대받고 있는 AV1 동영상 코덱 및 AVIF 코덱으로 빠르게 인코딩합니다. 반도체로 처리 속도를 가속화 한 덕분에 우수한 성능은 유지하면서 네트워크 사용량이 줄어 비용을 낮췄으며, 동일 화지 대비 기존 코덱(H.264, JPEG)보다 50% 적은 용량으로 저장 공간을 여유롭게 사용할 수 있습니다.

 

4. 4배 빠른 VMAF 측정 기술

최적의 화질을 찾아 인코딩 효율을 높이는 VAMF 스코어를 FPGA로 빠르게 측정해 CPU 대비 4배 빠르게 처리할 수 있습니다. 빠른 처리 속도로 네트워크 사용량을 줄여 효율성을 높였습니다.

Fast and Intelligent Video Semiconductor IP

 

1. AI based Video Enhancing and Upscaling

BLUEDOT's video AI improves the low-resolution video to high-resolution while enhancing video quality. By implementing video processing AI that requires numerous computations with semiconductor technology, video can be processed at high speed, and real-time video quality improvement and upscaling (SD to FHD) is possible.

 

2. Pre-processing video AI that keep perceptual quality but lower data 30%

Video Artificial Intelligence models identify pixels that don't affect image quality, so it can reduce data by 30%. Processes video in the pre-encoding stage to adapt to any codec, enabling 4Kp60 real-time processing on a single FPGA card.

 

3. High performance AV1 & AVIF codec

AV1 video codec and AVIF image codec, which are expected to be the next generation codecs with high compression efficiency than existing h.264 and jpeg, developed encoder by BLUEDOT to encoding faster and performance high. Accelerated encoding speeds with semiconductors have lowered costs by reducing network usage without sacrificing superior performance, and frees up storage space by using 50% less than traditional codecs (H.264, JPEG) for the same quality.

 

4. x4 Fast Process VMAF scoring

VMAF score, which finds the optimal video quality and improves encoding efficiency, FPGAs can measure x4 faster than CPUs. Increased efficiency by reducing network usage with faster processing speeds.