Emergent Vision Technologies eCapture Pro Software Now Supports GPUDirect H.265/H.264, and RTMP Streaming

Emergent Vision Technologies, announces that its eCapture Pro software now supports GPUDirect, H.265 and H.264 video codecs, and RTMP (real-time messaging protocol) streaming.

These additions further strengthen the award-winning software’s ability to enable the most challenging high-speed imaging and machine vision applications with no data loss.

Emergent’s eCapture Pro software offers an intuitive graphical interface for complete and easy system integration, camera setting flexibility, advanced preview and recording, and synchronized capture to microsecond accuracy, as well as calibration, background, and production-take management capabilities. In addition, the software enables the use of the industry’s fastest frame rates. With support for GPUDirect — a capability that passes data straight from a camera to a GPU, bypassing CPU and system memory — eCapture Pro can help remove the overhead and latency involved with making memory copies.

“GPUDirect significantly improves the speed of image acquisition for processing and video streaming, and/or image storage for high-speed imaging applications using eCapture Pro software,” said John Ilett, president and CTO at Emergent Vision Technologies. “With this feature, systems integrators and engineers do not need to worry about latency or jitter in mission-critical applications.”

High-Speed Compression and Streaming
Once data passes from the camera to a GPU, the now-supported H.265 and H.264 video codecs can compress the images. H.265, the successor to H.264, delivers up to 50% better video compression while maintaining the same level of video quality, making it an ideal option for high-speed, high-resolution video. Compatible with both H.265 and H.264, RTMP enables users to stream high-speed video to YouTube and other live streaming platforms.

In a real-world example involving H.265 recording and RTMP streaming, 24x 25GigE Bolt HB-12000-SB machine vision cameras featuring the 12MP Sony Pregius S IMX535 sensor were deployed along with a midrange server and CPU, 24-port switch, two dual-port 100GigE network interface cards (NICs), and two NVIDIA GPUs. Using GPUDirect technology and running at 60 fps, camera data is sent directly to the two GPUs, which perform H.265 compression. Compressed images are stored locally on a disk and streamed via RTMP to YouTube. Running at 30fps, the system can support as many as 48 cameras using a 48-port switch while still running on a single server and CPU.

“Using switches in such a setup allows customers to multiplex the cameras to fewer NIC ports while also proving 1 µs accuracy synchronization using IEEE 1588 PTP,” said Ilett. “Real-world examples such as this deliver zero data loss imaging technologies, which will become a requirement for the next phase of high-speed imaging.”

www.emergentvisiontec.com

  Ask For More Information…