From the Publisher: 
The accessible presentation of this book gives both a general view of the entire computer vision enterprise and also offers sufficient detail to be able to build useful applications. Users learn techniques that have proven to be useful by first-hand experience and a wide range of mathematical methods. A CD-ROM with every copy of the text contains source code for programming practice, color images, and illustrative movies. Comprehensive and up-to-date, this book includes essential topics that either reflect practical significance or are of theoretical importance. Topics are discussed in substantial and increasing depth. Application surveys describe numerous important application areas such as image based rendering and digital libraries. Many important algorithms broken down and illustrated in pseudo code. Appropriate for use by engineers as a comprehensive reference to the computer vision enterprise.

Computer vision : a modern approach = 计算机视觉 : 一种现代的方法

Icy is a collaborative platform for biological image analysis that extends reproducible research principles by facilitating and stimulating the contribution and sharing of algorithm-based tools and protocols between researchers. Current research in biology uses evermore complex computational and imaging tools. Here we describe Icy, a collaborative bioimage informatics platform that combines a community website for contributing and sharing tools and material, and software with a high-end visual programming framework for seamless development of sophisticated imaging workflows. Icy extends the reproducible research principles, by encouraging and facilitating the reusability, modularity, standardization and management of algorithms and protocols. Icy is free, open-source and available at http://icy.bioimageanalysis.org/
 .

Icy: an open bioimage informatics platform for extended reproducible research

Super-resolution reconstruction proposes a fusion of several low-quality images into one higher quality result with better optical resolution. Classic super-resolution techniques strongly rely on the availability of accurate motion estimation for this fusion task. When the motion is estimated inaccurately, as often happens for nonglobal motion fields, annoying artifacts appear in the super-resolved outcome. Encouraged by recent developments on the video denoising problem, where state-of-the-art algorithms are formed with no explicit motion estimation, we seek a super-resolution algorithm of similar nature that will allow processing sequences with general motion patterns. In this paper, we base our solution on the Nonlocal-Means (NLM) algorithm. We show how this denoising method is generalized to become a relatively simple super-resolution algorithm with no explicit motion estimation. Results on several test movies show that the proposed method is very successful in providing super-resolution on general sequences.

Generalizing the Nonlocal-Means to Super-Resolution Reconstruction

We must all accept that science is data and that data are science, and thus provide for, and justify the need for the support of, much-improved data curation. (Hanson, Sugden, & Alberts) 
 
Researchers are producing an unprecedented deluge of data by using new methods and instrumentation. Others may wish to mine these data for new discoveries and innovations. However, research data are not readily available as sharing is common in only a few fields such as astronomy and genomics. Data sharing practices in other fields vary widely. Moreover, research data take many forms, are handled in many ways, using many approaches, and often are difficult to interpret once removed from their initial context. Data sharing is thus a conundrum. Four rationales for sharing data are examined, drawing examples from the sciences, social sciences, and humanities: (1) to reproduce or to verify research, (2) to make results of publicly funded research available to the public, (3) to enable others to ask new questions of extant data, and (4) to advance the state of research and innovation. These rationales differ by the arguments for sharing, by beneficiaries, and by the motivations and incentives of the many stakeholders involved. The challenges are to understand which data might be shared, by whom, with whom, under what conditions, why, and to what effects. Answers will inform data policy and practice. © 2012 Wiley Periodicals, Inc.

科研数据共享的挑战 (The Conundrum of Sharing Research Data)

Super-resolution, the process of obtaining one or more high-resolution images from one or more low-resolution observations, has been a very attractive research topic over the last two decades. It has found practical applications in many real-world problems in different fields, from satellite and aerial imaging to medical image processing, to facial image analysis, text image analysis, sign and number plates reading, and biometrics recognition, to name a few. This has resulted in many research papers, each developing a new super-resolution algorithm for a specific purpose. The current comprehensive survey provides an overview of most of these published works by grouping them in a broad taxonomy. For each of the groups in the taxonomy, the basic concepts of the algorithms are first explained and then the paths through which each of these groups have evolved are given in detail, by mentioning the contributions of different authors to the basic concepts of each group. Furthermore, common issues in super-resolution algorithms, such as imaging models and registration algorithms, optimization of the cost functions employed, dealing with color information, improvement factors, assessment of super-resolution algorithms, and the most commonly employed databases are discussed.

/pdf/super-resolution-a-comprehensive-survey-xjl42yt9nt.pdf

Super-resolution: a comprehensive survey


Large acquisition times are required to achieve high-quality XCT scans, because of the need for high exposure times and a large number of X-ray projections. Reducing the number of projections results in an increase in noise, artefacts in the reconstruction domain and measurement errors. To enhance those low-quality XCT scans, while keeping acquisition times low, we investigated the possibilities of deep learning sinogram interpolation, by using a conditional generative adversarial network (cGAN) to artificially increase the number of X-ray projections before reconstruction. The method is evaluated on simulated XCT scans from real objects using the ASTRA toolbox. First, the hyperparameters of the loss functions were altered to determine the optimal combination. Second, we experimented with different acquisition schemes and finally the number of X-ray projections is gradually reduced to quantify the minimum required X-ray projections and acquisition time.


/pdf/deep-learning-based-sinogram-interpolation-applied-to-x-ray-3o6fkiy2.pdf

Deep learning based sinogram interpolation applied to X-ray CT measurements of polymer additive manufacturing parts

Recently, several works have proposed fusing radar data as an additional perceptual signal into monocular depth estimation models because radar data is robust against varying light and weather conditions. Although improved performances were reported in prior works, it is still hard to tell how much depth information radar can contribute to a depth estimation model. In this paper, we propose radar inference and supervision experiments to investigate the intrinsic depth potential of radar data using state-of-the-art depth estimation models on the nuScenes dataset. In the inference experiment, the model predicts depth by taking only radar as input to demonstrate the inference capability using radar data. In the supervision experiment, a monocular depth estimation model is trained under radar supervision to show the intrinsic depth information that radar can contribute. Our experiments demonstrate that the model using only sparse radar as input can detect the shape of surroundings to a certain extent in the predicted depth. Furthermore, the monocular depth estimation model supervised by preprocessed radar achieves a good performance compared to the baseline model trained with sparse lidar supervision.

How much depth information can radar contribute to a depth estimation model?

FIELD: calculating; counting.SUBSTANCE: invention relates to the computer equipment. Device for forming output three-dimensional image comprises: a first image forming module for forming an intermediate three-dimensional image, wherein is formed using the depth-based processing, number of image units of pixel values for plurality of regions, wherein number of image blocks is different for two areas from plurality of regions, and each image block contains pixel values for group of pixels corresponding to direction of view; a second image forming module for forming, using depth-based processing, an output three-dimensional image comprising a number of images of a view; an adaptation module for local adaptation of the number of image blocks formed by the first image forming module, with different viewing directions for the first region of the plurality of regions in response to the property of one of the intermediate image and the presentation of the three-dimensional scene, from which the first module generates an intermediate three-dimensional image.EFFECT: technical result is providing an improved three-dimensional image.14 cl, 11 dwg

Method and apparatus for three-dimensional image forming

Infrastructure maintenance of complex environments like railroads is a very expensive operation. Recent advances in mobile mapping systems to collect 3D point cloud data and in deep learning for detection and segmentation can prove to be very helpful in automating this maintenance and allowing preventive maintenance at certain locations before big failures occur. Some fully-supervised methods have been developed for understanding dynamic railroad environments. These methods often fail to generalize to infrastructure changes or new classes in low-labeled data. To address this issue, we propose a railroad segmentation method that leverages few-shot learning by generating class prototypes for the most relevant infrastructure classes. This method takes advantage of existing embedding networks for point clouds, taking the geometrical and spatial context into account for feature representation of complex connected classes. We evaluate our method on real-world data measured on Belgian railway tracks. Our model achieves promising results on connected classes, exposed to only a few annotated samples at test time. 

Few-shot learning on point clouds for railroad segmentation


The time of an X-ray Computed Tomography (XCT) measurement is directly affected by the number of acquired X-ray projections. In general, a large number of X-ray projections, hence a large acquisition time is required to obtain a qualitative reconstruction and successive feature analysis. To expand the applicability of XCT towards low-end parts we propose a novel sinogram interpolation method that incorporates the object rotation. The method combines a forward projection model of the XCT system and a deep learning regression model to generate intermediate X-ray projections for XCT scans with a reduced number of projections. Thereby, the accompanying reconstruction artefacts can be reduced while preserving a lower acquisition time. The method is validated on simulated projections of the 3D Shepp-Logan phantom and simulated projections of real 3D printed components using the ASTRA-toolbox. Within the reconstructions, less streaking artefacts are observed and an increased contrast between different features is obtained without introducing rotational artefacts.


Patrick Vandewalle

Papers

Deep learning based sinogram interpolation applied to X-ray CT measurements of polymer additive manufacturing parts

How much depth information can radar contribute to a depth estimation model?

Method and apparatus for three-dimensional image forming

Few-shot learning on point clouds for railroad segmentation

A machine learning supported sinogram interpolation method for X-ray computed tomography