The 3D reconstruction using deep learningbased intelligent systems can provide great help for measuring an individual’s height and shape quickly and accurately through 2D motionblurred images. Generally, during the acquisition of images in realtime, motion blur, caused by camera shaking or human motion, appears. Deep learningbased intelligent control applied in vision can help us solve the problem. To this end, we propose a 3D reconstruction method for motionblurred images using deep learning. First, we develop a BFWGAN algorithm that combines the bilateral filtering (BF) denoising theory with a Wasserstein generative adversarial network (WGAN) to remove motion blur. The bilateral filter denoising algorithm is used to remove the noise and to retain the details of the blurred image. Then, the blurred image and the corresponding sharp image are input into the WGAN. This algorithm distinguishes the motionblurred image from the corresponding sharp image according to the WGAN loss and perceptual loss functions. Next, we use the deblurred images generated by the BFWGAN algorithm for 3D reconstruction. We propose a threshold optimization random sample consensus (TORANSAC) algorithm that can remove the wrong relationship between two views in the 3D reconstructed model relatively accurately. Compared with the traditional RANSAC algorithm, the TORANSAC algorithm can adjust the threshold adaptively, which improves the accuracy of the 3D reconstruction results. The experimental results show that our BFWGAN algorithm has a better deblurring effect and higher efficiency than do other representative algorithms. In addition, the TORANSAC algorithm yields a calculation accuracy considerably higher than that of the traditional RANSAC algorithm.
Due to some factors, such as camera shaking and human motion, realtime image blurring easily occurs. For a good visual effect, it is very important to remove the blur and obtain a sharp image [
To solve the problems of the existing deep learning algorithms, we propose a BFWGAN algorithm, which combines the bilateral filtering (BF) [
3D reconstruction of the human body is very useful for the rapid and accurate measurement of an individual’s height and body shape [
For the 3D reconstruction of motionblurred images, we use the deblurred images generated with the BFWGAN algorithm to perform the 3D reconstruction. The most important part of 3D reconstruction is the calculation of the camera parameters, which mainly include the global rotation matrix and global translation vector for multiview 3D reconstruction [
The contributions of this paper are listed as follows:
We use deep learningbased intelligent systems to remove the motion blur in images. The BFWGAN algorithm is proposed, which combines the BF denoising theory with WGAN. The BF denoising algorithm is used to remove the noise and retain the details of the blurred image. The WGAN adopts the blurred image, and corresponding sharp images are input into the WGAN. The BFWGAN algorithm has a better deblurring effect and higher efficiency than other representative algorithms.
We adopt the deblurred images generated from the BFWGAN algorithm to perform the 3D reconstruction. The TORANSAC algorithm is proposed, which can remove the wrong relationship between two views in the 3D reconstructed model relatively accurately. Compared with the traditional RANSAC algorithm, the TORANSAC algorithm can adjust the threshold adaptively, which improves the accuracy of the 3D reconstruction results.
The remainder of this paper is organized as follows: Section 2 consists of two parts. Part 2.1 presents the deep learningbased intelligent systems to remove the motion blur of images through the BFWGAN algorithm, and Part 2.2 explains the TORANSAC algorithm that we used to perform the 3D reconstruction. In Section 3, we designed and evaluated an experiment to test the performance of the BFWGAN algorithm and the TORANSAC algorithm. In Section 4, we conclude our study and suggest directions for future work.
Normally, the processing of an image depends upon the quality, and the captured image in poor quality might result in a mistake. The intelligent systems using intelligent decisionmaking algorithms and techniques can help us to solve the image blurring problem.
In a mathematical model, image blurring can be described by the convolution process for an image. The original sharp image
where
A bilateral filter is a nonlinear denoising algorithm that eliminates noise while preserving image details [
Step 1: Compute the Gaussian weight region filter based on the spatial distances:
where
where
Step 2: Obtain the edge filter based on the degree of similarity:
where
where
Step 3: Create the bilateral filter by combining the Gaussian weight region filter with the edge filter:
where
After the local subregion
This paper proposes a WGAN deblurring algorithm that adopts both the WGAN loss and perceptual loss functions [
The WGAN between generator
where
① WGAN framework
As shown in
② Loss Function
The loss function of this paper consists of the WGAN loss and perceptual loss functions. The total loss function
where
where
where
Multiview 3D reconstruction is mainly composed of four parts: (1) Feature extraction and matching; (2) Camera parameter calculation; (3) 3D point cloud calculation; and (4) Bundle adjustment. The camera parameter calculation mainly involves the global rotation matrix and global translation vector for multiview 3D reconstruction [
The most commonly used method to calculate the global rotation matrix is the RANSAC algorithm [
The global rotation matrix is calculated by the relative rotation matrix through the leastsquares optimization algorithm. The formula is shown in
where
This paper proposes a TORANSAC algorithm to remove the wrong relationship between two views in the 3D reconstructed model. TORANSAC is a combination of the RANSAC algorithm and the threshold optimization concept. The use of different threshold parameters for the traditional RANSAC will affect the algorithm results. To avoid this problem, the TORANSAC algorithm is used to determine whether the model is reliable on the basis of the
where
The flow chart of the TORANSAC algorithm is shown in
Step 1: Determine the sampling times
where
Step 2: Calculate the initial global rotation matrix. Formula
Step 3: Calculate the errors for the remaining edges and sort the edges by the magnitude of the error. The error was calculated as the angle difference between the relative rotation matrix and the global rotation matrix, and the formula used is:
In formula
Step 4: Calculate the value of
Step 5: Select the edge set that minimizes the value of
For the performance evaluation of our approach, we collected 3000 realtime images of children from a kindergarten. There were 100 children aged 2–6 years, including 50 female students and 50 male students. A total of 30 realtime images were collected for each student in the JPG format. To evaluate the effect of the 3D reconstruction method for motionblurred images, we simulated the method in three parts. First, simulated noise images and blurred images were generated. The noise images and blurred images were generated by a ThinkPad S3490 computer [
The children were aged from 2–6 years, and one student of each age was selected as an example.
We chose the images of a 2yearold boy and a 4yearold girl to simulate the experiment. For the generation of simulated noise and blurred images, we mainly considered two aspects: the image noise parameters and motion blur parameters.
Gaussian noise is a common type of noise that occurs with camera shaking [
We used the MATLAB special function to blur the image and mainly considered two aspects: The blur angle and blur amplitude. For the blur angle, the blur amplitude was set to 15 pixels, and the blur angles studied were 30°, 45°, and 60°.
For comparison, we compared our algorithm with other image deblurring algorithms, including Xu L’s algorithm [
① Time Contrast Experiment
For the time contrast experiment of image deblurring, the images of a 2yearold boy and a 4yearold girl were selected. The experiment was repeated 3 times for each group, and then, the average value of three measurements was used for analysis.
Category  Time (s)  

Xu L’s algorithm  Chakrabarti A’s algorithm  Gong D’s algorithm  Nah S’s algorithm  BFWGAN 

2yearold boy  40.29  22.67  18.41  14.24  3.15 
4yearold girl  41.41  23.55  19.33  14.58  3.27 
② Accuracy Contrast Experiment
We adopt the peak signaltonoise ratio (PSNR) [
The TORANSAC and RANSAC algorithms were used to remove the wrong twoview relationships. For the 2yearold boy and 4yearold girl,
Algorithm  Number of removed wrong edges  Percentage of removed wrong edges (%) 

RANSAC  69  23.75 
ACRANSAC  15  5.25 
Algorithm  Number of removed wrong edges  Percentage of removed wrong edges (%) 

RANSAC  80  24.88 
ACRANSAC  19  5.93 
For the 2yearold boy and 4yearold girl,
Algorithm  Number of 3D points  Time (s) 

RANSAC  11452  11.24 
ACRANSAC  30841  12.15 
Algorithm  Number of 3D points  Time (s) 

RANSAC  12774  12.56 
ACRANSAC  34988  13.77 
The “intelligent” solutions are essential to take care of solving the blurring problem, which uses effective critical thinking procedures to restore the sharp image. First, we propose a BFWGAN algorithm to remove the motionblurred images, which combines the BF denoising theory with a WGAN. In this algorithm, the bilateral filter denoising algorithm is used to remove the noise and retain the details of the blurred image. Then, the blurred image and corresponding sharp image are input into the WGAN. This algorithm distinguishes the motionblurred image from the corresponding sharp image according to the WGAN loss and perceptual loss functions, which allows the fine texturerelated details to be revealed and the highprecision contours of the images to be revealed. Second, we used the deblurred images generated by the BFWGAN algorithm to perform 3D reconstruction. The TORANSAC algorithm is proposed, which can remove the wrong relationships between two views in the 3D reconstructed models relatively accurately. Compared with the traditional RANSAC algorithm, the TORANSAC algorithm can adjust the threshold adaptively, which improves the accuracy of the 3D reconstruction results. The experimental results show that our BFWGAN has a better deblurring effect and higher efficiency than do other representative algorithms. In addition, the TORANSAC 3D reconstruction algorithm yields a calculation accuracy considerably higher than that of the traditional RANSAC algorithm.
In a word, deep learning is significant for successfully executing image deblurring tasks. Effective deep learning algorithms can help yield more accurate 3D data, which can be used to measure individuals’ height and shape quickly and accurately. The vast use of these intelligent systems is due to its intelligent decisionmaking algorithms and techniques. However, deep learning trends in intelligent systems have the possibility of slowing down the entire computing process. There may be significant performance pressure on the processing and evaluation of images. In order to overcome these limitations in accuracy and computational time, we need to incorporate an effective deep learning image processing algorithm with an efficient data processing architecture in the future.
The author would like to thank the anonymous reviewers for their valuable comments and suggestions that improve the presentation of this paper.