Details preserved unsupervised depth estimation by fusing traditional stereo knowledge from laparoscopic images

Autor: Huoling Luo, Qingmao Hu, Fucang Jia
Jazyk: angličtina
Rok vydání: 2019
Předmět:
stereo image processing
surgery
image reconstruction
unsupervised learning
medical image processing
phantoms
image motion analysis
convolutional neural nets
traditional stereo method
proxy disparity labels
unreliable depth measurements
confidence measure
stereo accuracy
disparity images
rectified stereo images
proxy labels
smooth depth surface
unsupervised depth estimation
traditional stereo knowledge
laparoscopic images
vision-based laparoscope surgical navigation systems
truth depth
unsupervised learning depth estimation approach
dual encoder-decoder convolutional neural network
loss function
principled mask
parallax effects
neighbourhood smoothness term
constrain neighbouring pixels
partial nephrectomy da vinci surgery dataset
heart phantom data
hamlyn centre
Medical technology
R855-855.5
Zdroj: Healthcare Technology Letters (2019)
Druh dokumentu: article
ISSN: 2053-3713
DOI: 10.1049/htl.2019.0063
Popis: Depth estimation plays an important role in vision-based laparoscope surgical navigation systems. Most learning-based depth estimation methods require ground truth depth or disparity images for training; however, these data are difficult to obtain in laparoscopy. The authors present an unsupervised learning depth estimation approach by fusing traditional stereo knowledge. The traditional stereo method is used to generate proxy disparity labels, in which unreliable depth measurements are removed via a confidence measure to improve stereo accuracy. The disparity images are generated by training a dual encoder–decoder convolutional neural network from rectified stereo images coupled with proxy labels generated by the traditional stereo method. A principled mask is computed to exclude the pixels, which are not seen in one of views due to parallax effects from the calculation of loss function. Moreover, the neighbourhood smoothness term is employed to constrain neighbouring pixels with similar appearances to generate a smooth depth surface. This approach can make the depth of the projected point cloud closer to the real surgical site and preserve realistic details. The authors demonstrate the performance of the method by training and evaluation with a partial nephrectomy da Vinci surgery dataset and heart phantom data from the Hamlyn Centre.
Databáze: Directory of Open Access Journals
Externí odkaz: