Intelligent archive of laparotomic videos

本课题的研究目的是构建一个智能的剖腹手术视频档案。我们通过整合多个视频序列并利用图像处理技术删除外科医生的手和头部等障碍物，开发了自由视场高分辨率视频系统的方法。在自由视场高分辨率视频构建中，通过融合手术床周围多台摄像机的视频序列和医生头部可穿戴摄像机的视频序列，实现了高分辨率的自由视场视频。在移除障碍物中，一种识别外科医生的手和头并将其从融合的视频序列中删除以暴露手术视野的方法。通过人体模型的模拟手术室实验，验证了所提方法的有效性，并揭示了实现剖腹手术视频智能存档的问题。

The purpose of this research project is to construct the intelligent archive of laparotomic videos. We developed methods for the free-view high resolution video system by integrating several video sequences and deletion of obstructive objects such as surgeons hands and heads by utilizing the image processing techniques. In the free-view high resolution video construction, the free-view video with high resolution was achieved by fusing video sequences of several cameras around the surgical bed and wearable cameras on the surgeons heads. In the deletion of obstructive objects, a method for recognizing surgeons hands and heads and deleting them from a fused video sequence to expose the surgical field. As results of the experiment in the pseudo operating room using a human model, the efficacy of the proposed methods were confirmed and problems for achieving the intelligent archive of laparotomic videos were revealed.