基于可编程SOC的图像边缘检测系统

随着人工智能的不断发展，云计算面临数据量大、实时性低等问题。为了解决该问题，一种新的计算方式——边缘计算进入了人们的视野。边缘计算推动人工智能走向嵌入式系统，与传统嵌入系统相比，边缘人工智能要复杂得多，传统嵌入式设备计算能力弱、效率低的问题更为突出。基于Xilinx公司Zynq-7000系列的全可编程SOC完成系统设计，实现对视频的边缘检测。测试结果显示，硬件边缘检测速度比纯软件边缘检测速度快10倍，为边缘人工智能在嵌入式应用方面提供了可行方向。

Image Edge Detection System Based on Programmable SOC

With the continuous development of artificial intelligence， cloud computing is facing the problems of large amount of data and low real-time. To solve this problem， a new computing method， edge computing， has come into people’s vision. Edge computing drives AI to embedded systems. Compared with traditional embedded systems， edge AI is much more complex. Traditional embedded devices have weak computing power and low efficiency. In this paper， based on Xilinx Zynq-7000 series of fully programmable SOC to complete the system design， to achieve video edge detection. The test results show that the speed of hardware edge detection is 10 times faster than that of software edge detection， which provides a feasible direction for the application of edge artificial intelligence in embedded system.