Wyniki wyszukiwania - BazTech

1

Underwater image enhancement via efficient generative adversarial network

Qian Xin, Ge Peng

Optica Applicata

|

2021

|

Vol. 51, nr 4

483--497

EN

Underwater image enhancement has been receiving much attention due to its significance in facilitating various marine explorations. Inspired by the generative adversarial network (GAN) and residual network (ResNet) in many vision tasks, we propose a simplified designed ResNet model based on GAN called efficient GAN (EGAN) for underwater image enhancement. In particular, for the generator of EGAN we design a new pair of convolutional kernel size for the residual block in the ResNet. Secondly, we abandon batch normalization (BN) after every convolution layer for faster training and less artifacts. Finally, a smooth loss function is introduced for halo-effect alleviation. Extensive qualitative and quantitative experiments show that our methods accomplish considerable improvements compared to the state-of-the-art methods.

2

Enhanced depth of field and resolution three-dimensional integral imaging by variable spatial filtering and intermediate-view reconstruction technique

Huang Juezhen, Ge Peng

Optica Applicata

|

2021

|

Vol. 51, nr 4

551--563

EN

In this paper, we propose a novel three-dimensional (3D) integral imaging system to simultaneously improve the depth of field (DOF), resolution, and image quality of reconstructed images by variable spatial filtering and intermediate-view reconstruction technology (IVRT). In the proposed method, the camera performs element images acquisition on a 3D scene with objects of different depths through a 2D grid plane. The reconstructed slice image and block matching algorithm are used to extract the depth of the element images. To improve the sharpness of depth, the Laplace operator is used to perform variable depth filtering on objects of different depths, and depth-enhanced all-filtering element images are obtained through simple pixel fusion. IVRT is applied to all-filtering element images to obtain more element images to reconstruct a resolution-enhanced 3D image. According to the energy of gradient (EOG) value and the Tenengrad value, the reconstruction image quality evaluation of the proposed method is improved by 7.63 and 4.81 times compared with the traditional method, respectively. By the proposed method of generating all-filtering element images and an IVRT in 3D integral imaging system, the experimental results demonstrate that the 3D reconstructed image has extended depth of field, enhanced resolution and improved image quality.

3

Depth extraction in computational integral imaging based on bilinear interpolation

Huang Huihua, Ge Peng

Optica Applicata

|

2020

|

Vol. 50, nr 4

497--509

EN

We proposed a method using a merit function to determine the depth of objects in computational integral imaging by analyzing the existing methods for depth extraction of target objects. To improve the resolution of reconstructed slice images, we use a digital camera moving in horizontal and vertical direction with the set interval to get elemental images with high resolution and bilinear interpolation algorithm to increase the number of pixels in slice image which improves the resolution obviously. To show the feasibility of the proposed method, we carried out our experiment and presented the results. We also compared it with other merit functions. The results show that merit function SMD2 to determine the depth of objects is more accurate and suitable for real-time application.