This chapter described the use of Jetson TX1 in the deployment of CUDA and OpenCV code. The properties of the GPU device present on a TX1 board that make it ideal for deploying computationally complex applications are explained in detail. The performance of Jetson TX1 for CUDA applications such as adding two large arrays is measured and compared with GPUs present on laptops. The procedure to work with images on Jetson TX1 is explained in detail in this chapter. The image-processing applications like image addition, image thresholding, and image filtering are deployed on Jetson TX1 and performance is measured for them.
The best part of Jetson TX1 is that multiple cameras can be interfaced with it in an embedded environment, and videos from that camera can be processed to design complex computer vision applications. The procedure to capture video from an onboard or USB camera...