Records video clips of detected objects.Communicates over MQTT for easy integration into other systems.Object detection with TensorFlow runs in separate processes for maximum FPS.Uses a very low overhead motion detection to determine where to run object detection.Leverages multiprocessing heavily with an emphasis on realtime over processing every frame.Designed to minimize resource use and maximize performance by only looking for objects when and where it is necessary.Tight integration with Home Assistant via a custom component.According to the developer, Coral USB Accelerator can process 100+ FPS with very little overhead.Īs of this writing, Frigate has a few notable features: The Coral is simply a TPU coprocessor that will outperform even the best CPUs in video workloads. It uses OpenCV and Tensorflow to perform realtime object detection locally for IP cameras.įrigate also supports Google Coral Accelerator – the latest innovation in machine learning. Easy installation through unRAID Community Application platformįrigate is a newly developed local NVR which aims to power smart homes with tightly integrated0 AI object detection features.Masks to limit where motion detection occurs.Supports specifying zones to limit detection to a particular area to reduce false positives.Supports hardware acceleration on different platforms : CUDA for systems with a supported GPU, OpenCL, OpenMax and MMAL on the RaspberryPi 3B+, Intel QuickSync with VA-API.
Supports multiple different object detectors, record videos on detected objects:.Some of its notable features are listed below. However, it looks really promising as modern technologies are coming into the mix.įrom multiple object detection backend and extensive hardware acceleration support, Viseron may be the choice that suits you. Viseron is a very new open source NVR solutions and its features are being actively developed.
0 kommentar(er)
