Image focus volume regularization for shape from focus through 3D weighted least squares

Ali, Usman; Pruks, Vitalii; Mahmood, Muhammad Tariq

doi:10.1016/j.ins.2019.03.056

Cited by 21 publications

(13 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the weighted L1 (Kim et al, 2017) scheme, the same standard deviation volume u p was incorporated as the guidance volume. We also compared the proposed method with two related approaches (Ali et al, 2019; Ali & Mahmood, 2020) that have been proposed recently in SFF. These approaches minimize weighted L2-based energy function, and so does the proposed method.…”

Section: Resultsmentioning

confidence: 99%

“…However, the problem formulations in these three schemes are done in different domains and scenarios. Ali et al (2019) (WL2-FVR) proposed improvement of the initial focus volume (FV) by taking it as the fidelity term. This technique employs the image sequence as the guidance volume while doing regularization.…”

Section: Resultsmentioning

confidence: 99%

“…Jeon et al (2019) proposed an efficient FM operator that has a ring-and-disk-like structure and, thus, has properties of local and non-local FMs. Recently, Ali et al (2019) and Ali & Mahmood (2020) have proposed to improve the focus volumes through regularization. Mutahira et al (2020) employed a simple deep neural network (DNN) to aggregate 3D shape.…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Depth Hypotheses Fusion through 3D Weighted Least Squares in Shape from Focus

Ali

Mahmood

2021

Microsc Microanal

Self Cite

View full text Add to dashboard Cite

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Depth Hypotheses Fusion through 3D Weighted Least Squares in Shape from Focus

Ali

Mahmood

2021

Microsc Microanal

Self Cite

View full text Add to dashboard Cite

show abstract

“…Therefore, the core idea of SFF is to find the focus quality of each pixel in the image sequence and then estimate depth from the best focused pixels. Accordingly, at the second step of SFF, the focus quality of each pixel in the image sequence is computed by applying a focus measure (FM) operator and it provides an image focus volume (Ali, Pruks, & Mahmood, 2019). In the literature, a variety of FM operators have been suggested and each one exhibits its own strength and weakness.…”

Section: Introductionmentioning

confidence: 99%

Depth enhancement through correlation‐based guided filtering in shape from focus

Ali

Mahmood

2021

Microscopy Res & Technique

Self Cite

View full text Add to dashboard Cite

Mostly, shape‐from‐focus (SFF) techniques do not consider any guidance or prior information from the input image sequence while improving the depth information. Consequently, the resultant depth maps may have inaccuracies due to missing the fine structural details. In this paper, we propose a depth enhancement method based on a novel guidance map and guided filtering. In the proposed method, first, a focus measure is applied on image sequence (volume) to compute the focus volume and an initial depth map is obtained by maximizing the focus measure in the optical direction. The guidance map is computed based on the correlation among the image volume and the focus volume along the optical axis. Finally, the improved depth map is obtained by applying guided filtering of initial depth by incorporating the weights from the suggested guidance map. Experiments were carried out using synthetic and real and microscopic image sequences of various objects. Experimental results have demonstrated the effectiveness of the proposed method and a reasonable improvement has been achieved in the quality of reconstructed depth maps.

show abstract

“…Muhammad and Choi (2010) proposed the SFF method based on simple FM and Bezier Surface approximation. In order to increase the accuracy of 3D shapes, another method suggests applying 3D weighted least squares to enhance image focus volume (Ali, Pruks, and Mahmood (2019)). Several 3D shapes recovered by applying different FMs are combined into one final shape by utilizing statistical clue of the shape in (Ali and Mahmood (2019)).…”

Section: Introductionmentioning

confidence: 99%

A simplified approach using deep neural network for fast and accurate shape from focus

Mutahira

Muhammad

et al. 2020

Microscopy Res & Technique

View full text Add to dashboard Cite

Three-dimensional shape recovery is an important issue in the field of computer vision. Shape from Focus (SFF) is one of the passive techniques that uses focus information to estimate the three-dimensional shape of an object in the scene. Images are taken at multiple positions along the optical axis of the imaging device and are stored in a stack. In order to reconstruct the three dimensional shape of the object, the best-focused positions are acquired by maximizing the focus curves obtained via application of a focus measure operator. In this article, Deep Neural Network (DNN) is employed to extract the more accurate depth of each object point in the image stack. The size of each image in the stack is first reduced and then provided to the proposed DNN network to aggregate the shape. The initial shape is refined by applying a median filter, and later the reconstructed shape is sized back to original by utilizing bi-linear interpolation. The results are compared with commonly used focus measure operators by employing root mean squared error (RMSE), correlation, and image quality index (Q). Compared to other methods, the proposed SFF method using DNN shows higher precision and low computational time consumption.

show abstract

Image focus volume regularization for shape from focus through 3D weighted least squares

Cited by 21 publications

References 21 publications

Depth Hypotheses Fusion through 3D Weighted Least Squares in Shape from Focus

Depth Hypotheses Fusion through 3D Weighted Least Squares in Shape from Focus

Depth enhancement through correlation‐based guided filtering in shape from focus

A simplified approach using deep neural network for fast and accurate shape from focus

Contact Info

Product

Resources

About