Fast projective image rectification for planar objects with Manhattan structure

In this work, we present a network architecture with parallel convolutional neural networks (CNN) for removing perspective distortion in images. While other works generate corrected images through the use of generative adversarial networks or encoder-decoder networks, we propose a method wherein three CNNs are trained in parallel, to predict a certain element pair in the 3×3 transformation matrix, M^. The corrected image is produced by transforming the distorted input image using M^−1. The networks are trained from our generated distorted image dataset using KITTI images. Experimental results show promise in this approach, as our method is capable of correcting perspective distortions on images and outperforms other state-of-the-art methods. Our method also recovers the intended scale and proportion of the image, which is not observed in other works.

Section: Related Workmentioning

confidence: 99%

Blind First-Order Perspective Distortion Correction Using Parallel Convolutional Neural Networks

Gallego

Ilao

Cordel

2020

“…Note that the angular spectrum of the point source is uniform in 4π steradians or 2π radians in the 2D case and linear source. The part of this spectrum that participates in image formation is determined by the crystal diffraction reflection coefficient, which is the Fourier image of the Green function (24). Thus, the Green function describes the blurring of the point as a result of the limited perception of the input angular spectrum by the crystal analyzer.…”

Section: The Use Of a Crystal Analyzer In X-ray Imaging Schemesmentioning

confidence: 99%

“…where i is the number of projective base points. The transformation is performed over the coordinates of the projective basis points [22][23][24][25]. The lattice used as a calibration object does not limit the class of test objects because the theory of constructing projective-invariant bases for projective transformed smooth convex figures is also developed [26].…”

Section: Data Processing Algorithmmentioning

confidence: 99%

Enhanced Tomographic Sensing Multimodality with a Crystal Analyzer

Бузмаков

Chukalina

Dyachkova

et al. 2020

This article demonstrates how a combination of well-known tools—a standard 2D detector (CCD (charge-coupled device) camera) and a crystal analyzer—can improve the multimodality of X-ray imaging and tomographic sensing. The use of a crystal analyzer allowed two characteristic lines of the molybdenum anode—Kα and Kβ—to be separated from the polychromatic radiation of the conventional X-ray tube. Thus, as a result of one measurement, three radiographic projections (images) were simultaneously recorded. The projection images at different wavelengths were separated in space and registered independently for further processing, which is of interest for the spectral tomography method. A projective transformation to compensate for the geometric distortions that occur during asymmetric diffraction was used. The first experimental results presented here appear promising.

“…Detection of a chart adds the bounding box search to the classification task, as it needs to locate the charts on a document file, report, or scanned print. Some works have addressed the standard approach for perspective correction through image rectification while using vanishing points [13][14][15][16]. However, to the best of our knowledge, none applied to the scenario of chart or document images for chart recognition.…”

Section: Introductionmentioning

confidence: 99%

A Real-World Approach on the Problem of Chart Recognition Using Classification, Detection and Perspective Correction

Araújo

Chagas

Alves

et al. 2020

Data charts are widely used in our daily lives, being present in regular media, such as newspapers, magazines, web pages, books, and many others. In general, a well-constructed data chart leads to an intuitive understanding of its underlying data. In the same way, when data charts have wrong design choices, a redesign of these representations might be needed. However, in most cases, these charts are shown as a static image, which means that the original data are not usually available. Therefore, automatic methods could be applied to extract the underlying data from the chart images to allow these changes. The task of recognizing charts and extracting data from them is complex, largely due to the variety of chart types and their visual characteristics. Other features in real-world images that can make this task difficult are photo distortions, noise, alignment, etc. Two computer vision techniques that can assist this task and have been little explored in this context are perspective detection and correction. These methods transform a distorted and noisy chart in a clear chart, with its type ready for data extraction or other uses. This paper proposes a classification, detection, and perspective correction process that is suitable for real-world usage, when considering the data used for training a state-of-the-art model for the extraction of a chart in real-world photography. The results showed that, with slight changes, chart recognition methods are now ready for real-world charts, when taking time and accuracy into consideration.