iCatch’s SoCs are powering a diversity of camera applications in
iCatch Technology demonstrates Ci1/CR1 on Embedded Vision Summit
Hello everyone, iCatch Technology's latest CR1 automotive SoC is demonstrated at 2019 Embedded Vision Summit in Santa Clara, CA. CR1 leverages VeriSilicon's VIP Neural Network Processor to combine image processing and edge AI processing into one single chip.
iCatch Technology currently offers two automotive grade image processing solutions, Ci1 (ISP) and CR1 (SoC).
Ci1 is a sensing camera module ISP supporting up to 1080p60. It incorporates a hardware neural network processor unit (NPU), supporting multiple deep learning framework. (Caffe, TensorFlow, Yolo and more)
CR1 is a car recording SoC which includes all of Ci1’s features with the additional H.264 CODEC supporting up to 4Kp30 and a wide variety of peripheral interfaces.
In the demonstration, the platform is CR1A, running iCatch in-house developed “iRVC” software. “iRVC stands for intelligent rear view camera which can detects objects like Cars, Motorcycles, Bicycles, Pedestrians, Buses, Trucks.
The raw video file is input via HDMI to CR1A Single Board Camera (SBC) and the analyzed result by the embedded NPU is shown on the external monitor.
If you are interested in our solutions, please contact us via the following link.