In this paper, different regions of the collected data samples are grayed out by the image gray value analysis method to obtain the average gray value, and the evaluation levels obtained from the ALPI model classifica...
详细信息
In logging interpretation, electrical image data is used to identify important geological features such as fissures and cavities, providing crucial information for evaluating productivity. Therefore, obtaining high-qu...
详细信息
ISBN:
(纸本)9781959025436
In logging interpretation, electrical image data is used to identify important geological features such as fissures and cavities, providing crucial information for evaluating productivity. Therefore, obtaining high-quality, fully restored electrical image data is essential for both manual and AI-based image interpretation methods. However, traditional imageprocessingalgorithms often produce restored data with poor correlation between geological features across plates, as well as noise and jagged edges. To address this issue, we propose a reliable method for filling in missing sections of electrical image data that can predict high-quality images using deep learning algorithms. While traditional image inpainting algorithms fail to extract meaningful features from pad data, deep learning networks using the encoder-decoder architecture can automatically learn representations of borehole images. In this work, we propose a Generative Adversarial Network (GAN) based inpainting model that generates high-quality missing pad data with semantically meaningful features as a reference for image logging interpretation. We achieved a U-Net based model as an encoder to aggregate contextual information from various feature maps through borehole image data. Based on the encoder, a simple up-sampling decoder with convolution layers is applied to predict the data in the missing area of borehole image data. The pixel-wise discriminator is used to distinguish between real image data and the corresponding synthesized patch. Our proposed model outperforms traditional inpainting methods, achieving an average accuracy of over 89% when inpainting missing areas of real borehole image data, and further validating the superiority of our method through analyzing the image log data with different azimuths. As a result, the encoder in this work also demonstrates its ability to be an efficient backbone model for extracting significant information from borehole image data. Copyright 2025, Internat
Vision models excel in image classification but struggle to generalize to unseen data, such as classifying images from unseen domains or discovering novel categories. In this paper, we explore the relationship between...
In order to improve the segmentation accuracy of 3D point cloud model in feature ambiguous region, the unsupervised clustering algorithm based on surface fusion features combines the depth residuals with the normal de...
详细信息
Froth flotation is an important process in the mineral processing industry for extracting valuable materials. This work investigates online microscopic imaging and machine learning based image analysis methods for rea...
详细信息
image recovery algorithms applied to imaging systems such as lensless imaging in end devices are a high-speed edge computing process. Diffraction recovery algorithms typically include operations such as FFT/IFFT, ampl...
详细信息
The rapid development of AI technologies, such as machine learning and deep learning, has provided new directions for medical research. With extensive data mining and learning, AI has demonstrated strong potential in ...
详细信息
In the process of radar multi-target radar tracking, a variety of multiple targets will lead to a large amount of data to be processed, which increases the complexity of data processing and the computational load. At ...
详细信息
This paper aims to explore an innovative English defect reporting software that utilizes natural language processing (NLP) techniques, particularly hybrid neural networks and attention mechanisms, to improve the accur...
详细信息
In the field of medical imaging, C-arm systems play a pivotal role in surgeries, especially interventional surgeries. However, the current C-arm imaging system cannot adapt to the application scenarios due to algorith...
详细信息
暂无评论