Real-time endoscopy haze removal: a synthetical method
Multimedia Tools and Applications(2024)
摘要
In endoscopic surgeries, the smoke and haze generated by temperature difference and electrosurgical knife usage degrades the surgical view. While many researches on natural single image haze removal being proposed, few works are done for video haze removal of endoscopy view. In this paper, we proposed a synthetical method for endoscopy haze removal combining a Refined Dark Channel Prior (RDCP), Spatial-Temporal Markov Random Field (STMRF), Color Attenuation Prior (CAP) and parameter self-adaption. Qualitative and quantitative experiment results outperformed some traditional and deep learning based methods. Our proposed method is an effective and practical one.
更多查看译文
关键词
Video dehazing,Endoscopy haze removal,Refined Dark Channel prior,Markov random field,Color attenuation prior
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要