A generative adversarial network with adaptive constraints for multi-focus image fusion
- PDF / 1,923,408 Bytes
- 11 Pages / 595.276 x 790.866 pts Page_size
- 38 Downloads / 235 Views
(0123456789().,-volV)(0123456789(). ,- volV)
ORIGINAL ARTICLE
A generative adversarial network with adaptive constraints for multi-focus image fusion Jun Huang1 • Zhuliang Le1 • Yong Ma1 • Xiaoguang Mei1 • Fan Fan1 Received: 17 January 2020 / Accepted: 14 March 2020 Ó Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract In this paper, we propose a novel end-to-end model for multi-focus image fusion based on generative adversarial networks, termed as ACGAN. In our model, due to the different gradient distribution between the corresponding pixels of two source images, an adaptive weight block is proposed in our model to determine whether source pixels are focused or not based on the gradient. Under this guidance, we design a special loss function for forcing the fused image to have the same distribution as the focused regions in source images. In addition, a generator and a discriminator are trained to form a stable adversarial relationship. The generator is trained to generate a real-like fused image, which is expected to fool the discriminator. Correspondingly, the discriminator is trained to distinguish the generated fused image from the ground truth. Finally, the fused image is very close to ground truth in probability distribution. Qualitative and quantitative experiments are conducted on publicly available datasets, and the results demonstrate the superiority of our ACGAN over the state-ofthe-art, in terms of both visual effect and objective evaluation metrics. Keywords Multi-focus image fusion Adaptive weight block Generative adversarial networks End-to-end
1 Introduction Due to the limitations of optical lenses, it is often difficult for an imaging device to take an image in which all the objects are captured in focus [13]. Thus, only the objects within the depth-of-field (DOF) have sharp appearance in the photograph while other objects are likely to be blurred. Multi-focus image fusion is known as a valuable technique to obtain an all-in-focus image by fusing multiple images of the same scene taken with different focal settings, which & Fan Fan [email protected] Jun Huang [email protected] Zhuliang Le [email protected] Yong Ma [email protected] Xiaoguang Mei [email protected] 1
is beneficial for human or computer operators, and for further image-processing tasks, e.g., segmentation, feature extraction and object recognition [15, 18]. Therefore, multi-focus image fusion has become a significant research topic in the field of image processing [10]. In the past few decades, many methods for multi-focus image fusion are continually proposed by researchers, and these methods can be attributed to two categories: spatial domain methods and transform domain methods. The methods based on spatial domain can be further divided into three groups according to different fusion rules [2, 9, 10]: pixel-based, block-based, and region-based fusion methods. Among them, the activity level measurements generally adopt the gradient information as a reference. In terms of transform domain method
Data Loading...
