Computing inverse optical flow

Sánchez, Javier; Salgado, Agustín; Monzón, Nelson

doi:10.1016/j.patrec.2014.09.009

Cited by 8 publications

(3 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The method of Sánchez et al and the temporal extension of ROF are named as TCOF and ROF T . The standard datasets only provide the forward optical flows, so we calculate the backward motions using the method proposed in (Sánchez et al, 2015).…”

Section: Methodsmentioning

confidence: 99%

Comparison of the Optical Flow Quality for Video Denoising

Monzón

Sánchez

2020

Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Application

Self Cite

View full text Add to dashboard Cite

Video denoising techniques need to understand the motion present in the scenes. In the literature, many strategies guide their temporal filters according to trajectories controlled by optical flow. However, the quality of these flows is rarely investigated. In fact, there are very few studies that compare the behavior of denoising proposals with different optical flow algorithms. In that direction, we analyze several methods and their performance using a general pipeline that reduces the noise through an average of the pixel's trajectories. This ensures that the denoising strongly depends on the optical flow. We also analyze the behavior of the methods at occlusions and illumination changes. The pipeline incorporates a process to get rid of these effects, so that they do not affect the comparison metrics. We are led to propose a ranking of optical flows methods depending on their efficiency for video denoising, that mainly depends on their complexity. 2 DENOISING FRAMEWORK Next, we briefly describe the main features of the denoising framework used to compare the influence of the optical flow methods in noise reduction.

show abstract

Section: Methodsmentioning

confidence: 99%

Comparison of the Optical Flow Quality for Video Denoising

Monzón

Sánchez

2020

Proceedings of the 15th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Application

Self Cite

View full text Add to dashboard Cite

show abstract

“…We used an interpolation algorithm to find V Ã , as described by Sanchez et al [14], and extended the methodology to three dimensional datasets. In fact, V Ã x ð Þ maps the voxels in the second frame to the voxels in the first frame.…”

Section: Methodsmentioning

confidence: 99%

A framework for synthetic validation of 3D echocardiographic particle image velocimetry

Falahatpisheh

Kheradvar

2015

Meccanica

View full text Add to dashboard Cite

Particle image velocimetry (PIV) has been significantly advanced since its conception in early 1990s. With the advancement of imaging modalities, applications of 2D PIV have far expanded into biology and medicine. One example is echocardiographic particle image velocimetry that is used for in vivo mapping of the flow inside the heart chambers with opaque boundaries. Velocimetry methods can help better understanding the biomechanical problems. The current trend is to develop three-dimensional velocimetry techniques that take advantage of modern medical imaging tools. This study provides a novel framework for validation of velocimetry methods that are inherently three dimensional such as but not limited to those acquired by 3D echocardiography machines. This framework creates 3D synthetic fields based on a known 3D velocity field V and a given 3D brightness field B. The method begins with computing the inverse flow V Ã based on the velocity field V. Then the transformation of B, imposed by V, is calculated using the computed inverse flow according toÞ , where x is the coordinates of voxels in B Ã , with a 3D weighted average interpolation, which provides high accuracy, low memory requirement, and low computational time. To check the validity of the framework, we generate pairs of 3D brightness fields by employing Hill's spherical vortex velocity field. B and the generated B Ã are then processed by our inhouse 3D particle image velocimetry software to obtain the interrelated velocity field. The results indicates that the computed and imposed velocity fields are in agreement.

show abstract

“…Estimation of the optical flow is defined to be backwards in time, from x i+1 , to x i . Defining this is important because the optical flow estimation is not symmetric due to effects such as occluded regions [41].…”

Section: A2 Usage Of the Temporal Informationmentioning

confidence: 99%

How can we learn (more) from challenges? A statistical approach to driving future algorithm development

Ross¹,

Bruno²,

Reinke³

et al. 2021

Preprint

View full text Add to dashboard Cite

Challenges have become the state-of-the-art approach to benchmark image analysis algorithms in a comparative manner. While the validation on identical data sets was a great step forward, results analysis is often restricted to pure ranking tables, leaving relevant questions unanswered. Specifically, little effort has been put into the systematic investigation on what characterizes images in which state-of-the-art algorithms fail. To address this gap in the literature, we (1) present a statistical framework for learning from challenges and (2) instantiate it for the specific task of instrument instance segmentation in laparoscopic videos. Our framework relies on the semantic meta data annotation of images, which serves as foundation for a General Linear Mixed Models (GLMM) analysis. Based on 51,542 meta data annotations performed on 2,728 images, we applied our approach to the results of the Robust Medical Instrument Segmentation Challenge (ROBUST-MIS) challenge 2019 and revealed underexposure, motion and occlusion of instruments as well as the presence of smoke or other objects in the background as major sources of algorithm failure. Our subsequent method development, tailored to the specific remaining issues, yielded a deep learning model with state-of-the-art overall performance and specific strengths in the processing of images in which previous methods tended to fail. Due to the objectivity and generic applicability of our approach, it could become a valuable tool for validation in the field of medical image analysis and beyond. and segmentation of small, crossing, moving and transparent instrument(s) (parts). Keywords surgical data science • image characteristics driven algorithm development • minimally invasive surgery • endoscopic vision • grand challenges • biomedical image analysis challenges • generalized linear mixed models • instrument segmentation • deep learning • artificial intelligence

show abstract

Computing inverse optical flow

Cited by 8 publications

References 22 publications

Comparison of the Optical Flow Quality for Video Denoising

Comparison of the Optical Flow Quality for Video Denoising

A framework for synthetic validation of 3D echocardiographic particle image velocimetry

How can we learn (more) from challenges? A statistical approach to driving future algorithm development

Contact Info

Product

Resources

About