Motion Detection and High Speed Camera Image Trigger

Detecting a slightest motion in the camera’s field of view and starting video recording is a common task in many surveillance and security systems. The problem is to detect the motion fast enough to be able to capture it in frame, so that the camera wouldn’t record empty video.

The software or FPGA detects a moving object by building histograms of input frames and comparing adjacent frames’ histograms for differences. If a histogram changes significantly, this signalizes about a new object moving in the shot. Such process is called Image Trigger. Image trigger works as a signal to begin capturing of several or multiple frames from a high-speed camera until the movement stops or another trigger activates.

Fastvideo developed such image trigger as PC software. We believe that CPU computations don’t allow attaining the desirable productivity and FPGA solutions are too complicated, so we transferred the histogram calculation to GPU and applied NVIDIA CUDA parallel processing technology. We use CameraLink high speed cameras with PCI-Express framegrabber as a data source for imaging system. Camera sends all data to PC RAM and our software is capable to calculate histogram for every image received in real time. Such architecture still has some latency, but the problem could be easily solved by organizing a circle buffer in PC RAM. Upon every trigger event software calculates a histogram data set and determines whether it differs from the starting histogram. Unprecedented performance of our Image Trigger technology opens a wide range of applications.

GPU image trigger features

• Data input: uncompressed 8-bit grayscale image
• Data output: final 256-bin histogram
• Continuous data mode (input one image after another)
• Parallel performance for image loading and computations
• Unlimited input image size
• Compatibility with NVIDIA mobile and table top video cards

PC test configuration

• ASUS P6T Deluxe V2 LGA1355, Core i7 2600, 3.4 GHz, DDR-III 8 GB
• GPU GeForce GTX 980
• OS Windows-10, 64-bit, CUDA 10

Applications for high-performance histogram calculations

High-performance CUDA histogram is suitable for a variety of digital imaging applications

• Image trigger for high speed cameras
• Image processing

What we could offer for your NVIDIA GPU

• CUDA histogram calculations for high speed imaging
• HPC custom solutions for CUDA image processing tasks
• Custom software development for NVIDIA GPU

Support

• Full technical support up to successful client integration

For any further information on these solutions please contact us via email.