基于i.MX6Q平台的图像融合实现

Implementation of image fusion based on i.MX6Q platform

为了实现手持式电晕移动检测设备的技术验证,选用NXP公司低功耗、高性能的i.MX6Q嵌入式开发平台进行双路视频图像的采集和融合。文中构建以视频图像处理单元(VPU)、图像处理单元(IPU)以及基于Linux V4L2架构的摄像头为主的驱动框架,并在内核驱动程序的基础上,分别对硬件抽象层、应用程序框架层以及应用程序层进行访问接口与控制算法的程序设计。最终通过对双路摄像头采集的图像进行Alpha blending测试结果表明,人机交互界面友好,视频采集图像稳定清晰,融合运行效果流畅,为后续紫外可见光双路探测的研究奠定了基础。

The low.power and high.performance i.MX6Q embedded development platform of NXP is selected to collect and fuse dual.channel video images,so as to realize technology verification of the hand.held corona mobile detector.The drive frame.work mainly constituted by the video processing unit(VPU),image processing unit(IPU)and camera based on the Linux V4L2 architecture is constructed in this paper.The program design of the access interface and control algorithm is conducted for the hardware abstraction layer,application program framework layer,and application program layer on the basis of the kernel drive program.The images collected by the dual.channel camera are processed by the Alpha blending algorithm,so as to realize image fusion of dual.channel visible light videos.The test results show that the man.machine interaction interface is friendly,the video acquisition images are stable and clear,and the fusion operation effect is smooth,which lays a foundation for the subse.quent study of UV.visible light dual.channel detection.