摘要
目的针对由于图像数据量巨大造成的传输延时过大问题设计一种远程区域图像传输算法。方法利用手术端对感兴趣区域进行检测,通过对内镜视频中的模糊区域、亮区和暗区进行剔除得到只包含关注信息的掩膜帧;远程传输采用模块化方法实现视频编码、码流解码和网络收发程序;操作端使用Criminisi算法对ROI区域进行掩膜修复,提高视频质量。结果只传输内镜图像的关注信息,平均可降低8.3%的数据量;采用Criminisi算法修复掩膜图像可提高清晰度,整个图像优化过程用时约为93 ms。结论该远程区域传输算法具有延时低、快速和高效等特点,可顺利完成消化内镜远程诊疗,有效提高偏远地区的医疗水平。
Objective A remote partial image transmission algorithm is designed to solve the problem of transmission delay caused by huge amount of image data.Methods The surgical terminal detected the region of interest(ROI),and removed the fuzzy area,bright area and dark area in the endoscopic video,then obtained a mask frame containing only the information of interest;The remote transmission used a modular method to realize video encoding,code stream decoding and network interactive programming.The operation terminal used the Criminisi algorithm to repair the ROI area to improve video quality.Results Only the ROI of endoscopic images was transmitted,which could reduce the data volume by 8.3%on average.The Criminisi algorithm could repair the mask image,and the whole image optimization process took about 93 ms.Conclusion The remote partial transmission algorithm has the characteristics of low delay,high speediness and high efficiency.It can successfully complete the remote diagnosis and treatment of digestive endoscopy and effectively improve the medical level in remote areas.
作者
张静
ZHANG Jing(Department of Military Biomedical Engineering,Air Force Medical University,Xi’an Shaanxi 710032,China)
出处
《中国医疗设备》
2021年第5期42-46,共5页
China Medical Devices
基金
军队重点项目(BWS14J038)。
