Asynchronous hardware decoder NGP

The asynchronous hardware decoder NGP (GPU+CPU) is a hardware and software algorithm for smart video surveillance systems that processes high-resolution video streams. The technology shifts decoding tasks from the CPU to the GPU, processing data nonlinearly. The system dynamically distributes the workload: the decoder simultaneously uses the CPU core and the GPU core to process complex codecs.

Key features

• Reduced CPU load: the CPU is freed from resource-intensive video decompression, improving system stability and allowing it to handle a larger number of cameras.

• Increased performance: the asynchronous method decodes frames as they are received, avoiding display delays (lags).

• Software specifics: typically used in video analytics and recognition modules where speed is critical, such as when detecting speeding violations in a video stream or identifying faces in a crowd.

Application

• Smart transport systems (see Configuration of speed estimation by video).
• Automatic speed control systems.
• License plate and face recognition systems (see Configuration of recognition parameters of captured faces).