Simple Scenes Research Articles

Pairwise Coarse Registration of Indoor Point Clouds Using 2D Line Features

Registration is essential for terrestrial LiDAR (light detection and ranging) scanning point clouds. The registration of indoor point clouds is especially challenging due to the occlusion and self-similarity of indoor structures. This paper proposes a 4 degrees of freedom (4DOF) coarse registration method that fully takes advantage of the knowledge that the equipment is levelled or the inclination compensated for by a tilt sensor in data acquisition. The method decomposes the 4DOF registration problem into two parts: (1) horizontal alignment using ortho-projected images and (2) vertical alignment. The ortho-projected images are generated using points between the floor and ceiling, and the horizontal alignment is achieved by the matching of the source and target ortho-projected images using the 2D line features detected from them. The vertical alignment is achieved by making the height of the floor and ceiling in the source and target points equivalent. Two datasets, one with five stations and the other with 20 stations, were used to evaluate the performance of the proposed method. The experimental results showed that the proposed method achieved 80% and 63% successful registration rates (SRRs) in a simple scene and a challenging scene, respectively. The SRR in the simple scene is only lower than that of the keypoint-based four-point congruent set (K4PCS) method. The SRR in the challenging scene is better than all five comparison methods. Even though the proposed method still has some limitations, the proposed method provides an alternative to solve the indoor point cloud registration problem.

A novel algorithm to improve image reconstruction quality for 2D streak camera

The single shot compressed ultrafast photography (CUP) camera is the fastest camera available. This technique is based on random codes to encode and reconstruct the ultrafast dynamical scene by a compressed sensing algorithm. In this paper, a new 2D ultrafast camera with high reliability was introduced and a novel reconstruction algorithm named GAP-TV was proposed. Firstly, this camera was simulated with the TwIST and GAP-TV methods for a simple scene; the result showed that the GAP-TV method had better PSNR and SSIM performance than TwIST. Secondly, the optical propagation process data was performed to this proposed 2D camera, and the result showed that this novel method can reconstruct it well. In addition, the efficiency of the GAP-TV method is excellent than TwIST. Finally, an experiment was conducted to capture the transient event when a femtosecond laser pulse passed through a beam splitter. The result showed that the GAP-TV method can reconstruct the compressed 2D image well, and the CUP system can achieve 1011 frames per second. So, a novel 2D ultrafast camera with high temporal resolution (4 picosecond) was designed and verified. It has wide applications in inertial confined fusion, fluorescence lifetime imaging, and nuclear physics.

Hybrid Transfer Learning and Broad Learning System for Wearing Mask Detection in the COVID-19 Era.

In the era of Corona Virus Disease 2019 (COVID-19), wearing a mask can effectively protect people from infection risk and largely decrease the spread in public places, such as hospitals and airports. This brings a demand for the monitoring instruments that are required to detect people who are wearing masks. However, this is not the objective of existing face detection algorithms. In this article, we propose a two-stage approach to detect wearing masks using hybrid machine learning techniques. The first stage is designed to detect candidate wearing mask regions as many as possible, which is based on the transfer model of Faster_RCNN and InceptionV2 structure, while the second stage is designed to verify the real facial masks using a broad learning system. It is implemented by training a two-class model. Moreover, this article proposes a data set for wearing mask detection (WMD) that includes 7804 realistic images. The data set has 26403 wearing masks and covers multiple scenes, which is available at "https://github.com/BingshuCV/WMD." Experiments conducted on the data set demonstrate that the proposed approach achieves an overall accuracy of 97.32% for simple scene and an overall accuracy of 91.13% for the complex scene, outperforming the compared methods.

Masked Face Detection Algorithm in the Dense Crowd Based on Federated Learning

Wearing masks is an effective and simple method to prevent the spread of the COVID‐19 pandemic in public places, such as train stations, classrooms, and streets. It is of positive significance to urge people to wear masks with computer vision technology. However, the existing detection methods are mainly for simple scenes, and facial missing detection is prone to occur in dense crowds with different scales and occlusions. Moreover, the data obtained by surveillance cameras in public places are difficult to be collected for centralized training, due to the privacy of individuals. In order to solve these problems, a cascaded network is proposed: the first level is the Dilation RetinaNet Face Location (DRFL) Network, which contains Enhanced Receptive Field Context (ERFC) module with the dilation convolution, aiming to reduce network parameters and locate faces of different scales. In order to adapt to embedded camera devices, the second level is the SRNet20 network, which is created by Neural Architecture Search (NAS). Due to privacy protection, it is difficult for surveillance video to share in practice, so our SRNet20 network is trained in federated learning. Meanwhile, we have made a masked face dataset containing about 20,000 images. Finally, the experiments highlight that the detection mAP of the face location is 90.6% on the Wider Face dataset, and the classification mAP of the masked face classification is 98.5% on the dataset we made, which means our cascaded network can detect masked faces in dense crowd scenes well.

The Rock from the Sky by Jon Klassen

Reviewed by: The Rock from the Sky by Jon Klassen Deborah Stevenson, Editor Klassen, Jon The Rock from the Sky; written and illus. by Jon Klassen. Candlewick, 2021 [96p] Trade ed. ISBN 9781536215625 $18.99 E-book ed. ISBN 9781536222685 $18.99 Reviewed from digital galleys R* Gr. 1-3 In this spare, humorous early chapter book of five slim chapters, the irresistible force looks set to meet the immovable object as a huge boulder plunges from the sky, hurtling toward an oblivious turtle, who muses happily about standing in its favorite spot. A groundhog-like critter has a “bad feeling” about that spot, and eventually, in the name of togetherness, the turtle moves to join the groundhog and its friend the snake—in the nick of time. Other chapters involve the turtle’s falling off of the now-landed rock (“Do you need help?” “No. . . . I never need help”), the animals’ dreaming of the future, a not-quite-shared sunset watching experience, and a terrapin sulk about the cozy nap the other two enjoy (“I see. I see how it is. Just enough room for two”). There’s rich humor in the clipped dialogue (differentiated by font color, but in main text rather than speech bubbles and a cool gray vs. black usage that’s a world away from Elephant & Piggie) that reveals our starring turtle in a few master strokes as a stubborn reptilian Eeyore type, and the comedy of terse dialogue leans magisterially into Beckett as performed by Buster Keaton. The brilliant coup de grâce that will slay audiences is the silent arrival of a huge weird creature that looks like a pony-sized six-legged spider with an eyeball for a body, which proves to be able to zap foliage into ashes, and which plays an abrupt and hilarious role in the book’s conclusion. The art is fascinatingly cinematic in a flip-book kind of way, using the static nature of the simple scenes and repeated compositions to set off the drama of small gestures, changes of position, and eye shifts; of course, this being Klassen, all three animals sport natty headgear, ranging from bijou bowlers to a lovely little beret for the snake. The sly humor and touch of sci-fi will make this a winner for reluctant later-stage readers as well as beginning readers developing their stamina, and it’ll be a great conspiratorial co-read with adults or sibs. Copyright © 2021 The Board of Trustees of the University of Illinois

A Bit on the Right, A Bit on the Left: Towards Logical Bundle Adjustment

This article proposes an innovative approach fully based on logic to determine the relative positions and orientations of objects in a scene photographed from different points of view as well as those of the cameras used to take the pictures. The proposal is absolutely not based on 2D feature extraction, projective geometry or least squares adjustment but on a logical approach based on an enumeration of simple relationships between the objects visible in the photos. It is an approach imitating a natural and unconscious reasoning that each of us makes by observing a scene: is this object more to the right than this one? And is this other one further away from me than the one who’s partially hiding it from me? It is therefore a question of approaching the problem by identifying and recognizing objects in photographs and not by measuring millions of points in space without having any idea of the object to which they belong. This article presents a ”proof of concept” based on virtual experimentation: in a discrete 3D space, a simple scene, composed of spheres of different colors and cameras, is modelled in a 3D format. In this work the positioning of the spheres and cameras is limited to a plane. Cameras are placed in the scene in order to see the spheres and then for each camera an image is generated. The application reads each image and deducts relationships between object and camera. These relationships based on the visible occlusions between the projections of the objects onto the photographs, are formalized according to Allen’s relationships. A knowledge base is implemented to allow an iterative process of SPARQL queries for qualitative spatial reasoning leading to a set of possible solutions. Finally, the system deduces the relative positions between objects and cameras and the result is imported and can be used within several photogrammetry software suites.

Radiative backpropagation

Physically based differentiable rendering has recently evolved into a powerful tool for solving inverse problems involving light. Methods in this area perform a differentiable simulation of the physical process of light transport and scattering to estimate partial derivatives relating scene parameters to pixels in the rendered image. Together with gradient-based optimization, such algorithms have interesting applications in diverse disciplines, e.g., to improve the reconstruction of 3D scenes, while accounting for interreflection and transparency, or to design meta-materials with specified optical properties. The most versatile differentiable rendering algorithms rely on reverse-mode differentiation to compute all requested derivatives at once, enabling optimization of scene descriptions with millions of free parameters. However, a severe limitation of the reverse-mode approach is that it requires a detailed transcript of the computation that is subsequently replayed to back-propagate derivatives to the scene parameters. The transcript of typical renderings is extremely large, exceeding the available system memory by many orders of magnitude, hence current methods are limited to simple scenes rendered at low resolutions and sample counts. We introduce radiative backpropagation , a fundamentally different approach to differentiable rendering that does not require a transcript, greatly improving its scalability and efficiency. Our main insight is that reverse-mode propagation through a rendering algorithm can be interpreted as the solution of a continuous transport problem involving the partial derivative of radiance with respect to the optimization objective. This quantity is "emitted" by sensors, "scattered" by the scene, and eventually "received" by objects with differentiable parameters. Differentiable rendering then decomposes into two separate primal and adjoint simulation steps that scale to complex scenes rendered at high resolutions. We also investigated biased variants of this algorithm and find that they considerably improve both runtime and convergence speed. We showcase an efficient GPU implementation of radiative backpropagation and compare its performance and the quality of its gradients to prior work.

Modeling standard plenoptic camera by an equivalent camera array

Postprocessing of light fields enables us to extract more information from a scene compared to traditional cameras. Plenoptic cameras and camera arrays are two common methods for light field capture. It has been long recognized that the two devices are in some ways equivalent. Practically, both techniques have important constraints. Camera arrays are unable to provide high angular sampling, and the plenoptic camera can have a limited spatial sampling. In simulation, we can easily explore both constraints by simulating two-dimensional viewpoint images and combining them into a four-dimensional light field. We present a transformation for converting between equivalent plenoptic configurations and camera arrays when they capture pristine light fields produced in simulation. We use this approach to simulate light fields of simple scenes and validate our transformation by comparing the focus distance of a standard plenoptic camera and the equivalent camera array’s light field. We also show how some simple practical effects can be added to the pristine, synthetic light field via postprocessing and their effect on refocusing distance.

Big Data Driven Detection of Trees in Suburban Scenes Using Visual Spectrum Eye Level Photography.

The aim of the work described in this paper is to detect trees in eye level view images. Unlike previous work that universally considers highly constrained environments, such as natural parks and wooded areas, or simple scenes with little clutter and clear tree separation, our focus is on much more challenging suburban scenes, which are rich in clutter and highly variable in type and appearance (houses, falls, shrubs, cars, bicycles, pedestrians, hydrants, lamp posts, etc.). Thus, we motivate and introduce three different approaches: (i) a conventional computer vision based approach, employing manually engineered steps and making use of explicit human knowledge of the application domain, (ii) a more machine learning oriented approach, which learns from densely extracted local features in the form of scale invariant features (SIFT), and (iii) a machine learning based approach, which employs both colour and appearance models as a means of making the most of available discriminative information. We also make a significant contribution in regards to the collection of training and evaluation data. In contrast to the existing work, which relies on manual data collection (thus risking unintended bias) or corpora constrained in variability and limited in size (thus not allowing for reliable generalisation inferences to be made), we show how large amounts of representative data can be collected automatically using freely available tools, such as Google’s Street View, and equally automatically processed to produce a large corpus of minimally biased imagery. Using a large data set collected in the manner and comprising tens of thousands of images, we confirm our theoretical arguments that motivated our machine learning based and colour-aware histograms of oriented gradients based method, which achieved a recall of 95% and precision of 97%.

Learners’ harmonic preferences are modulated by lexical retrieval difficulty

Typological work has established the existence of language universals – features or combinations of features that (co-) occur in unrelated languages more frequently than expected by chance. The origins of language universals are a fundamental question in language sciences as these universals are considered a reflection of cognitive mechanisms underlying human language. In this study, we use a miniature artificial language learning paradigm to explore whether a well-known universal – a preference for harmonic word ordering between adpositional and verb phrases (i.e., placing the head in a consistent position with respect to the complements across the two phrase types) – originates in biases during language learning and whether this preference interacts with memory constraints (operationalized as lexical retrieval difficulty). We first trained participants on miniature languages containing adpositional phrases of a deterministic word order (either prepositional or postpositional inputs) and then briefly exposed them to simple transitive sentences (verb-subject-object and subject-object-verb order, equally frequent in the input). At test, we asked learners to describe simple transitive scenes. We found that in the hard lexical retrieval condition, learners exposed to a postpositional language showed a preference for harmonic ordering but learners of the prepositional language did not, which is only partially consistent to the typological distribution. In the easy lexical retrieval condition, learners of neither the postpositional nor prepositional language showed a preference for harmonic ordering, indicating that this preference is modulated by memory constraints.

Survey of Research Progress on Target Detection and Discrimination of Single-channel SAR Images for Complex Scenes

As an active microwave imaging sensor, Synthetic Aperture Radar (SAR) has become one of the main means of Earth observation owing to its unique technical advantages of all-day, all-weather operation and long working distance. As such, it plays a very important role in military and civilian fields. With the development of SAR remote-sensing technology, high-resolution, high-quality SAR images are produced continuously. However, manual detection and recognition of targets of interest is time-consuming and laborious, so the development of Automatic Target Recognition (ATR) technology is a matter of urgency. The typical SAR ATR system primarily comprises three stages: detection, discrimination, and classification/recognition. The detection and discrimination stages are the basis of the SAR ATR system, and research on SAR applications in the radar field has been conducted by researchers around the world. For single-channel SAR images, target detection and discrimination from simple scenes yield good results. However, in complex scenes, the clutter scattering intensity is relatively high, the clutter background is heterogenous, the target scattering intensity is relatively weak, and the target distribution is dense. These factors continue to make accurate SAR target detection and discrimination difficult. In this paper, we summarize the recent research progress on single-channel SAR target detection and discrimination methods for complex scenes, analyze the characteristics and problems associated with various methods, and consider the future development trend of single-channel SAR target detection and discrimination methods for complex scenes.

Semantic segmentation of very high-resolution remote sensing image based on multiple band combinations and patchwise scene analysis

Large intraclass variance and low interclass variance are among the most challenging problems in very high-resolution (VHR) image classification. Semantic segmentation constructed in a deep convolution neural network is used as a classification algorithm conducted via end-to-end training, which combines spectral–spatial features and context information. However, large-scale remote sensing images cannot be directly processed because they are limited by GPU memory and segmentation algorithm. At the same time, classification using single band combinations is also unsatisfactory due to the extraordinary complex features of VHR images. Therefore, a method is proposed based on multiple band combinations and patchwise scene analysis. A complex remote sensing image can be considered as into a combination of simple scenes from multiple patchwise images. And optimal band combinations of each patchwise image are selected according to their scene. The segmentation results of each patchwise image are merged to get the desired results according to geographical coordinates. Our method is validated on the ISPRS 2-D Semantic Labeling dataset of Potsdam, on which results competitive with the state-of-the-art are obtained. The proposed scheme has strong universality and can be used for large-scale high-resolution remote sensing image classification.

Pattern recognition of decorative elements based on neural network

The Covid-19 first occurs in Wuhan, China in December 2019. After that, the virus has spread all over the world and at the time of writing this paper the total number of confirmed cases are above 11 million with over 600,000 deaths. The pattern recognition of complex environment can be used to determine if a COVID-19 breath pattern can be established with accuracy. The traditional decorative pattern detection method has a high degree of recognition in simple scene. However, the efficiency of decorative pattern detection in complex scenes is low and the recognition accuracy is not high. Firstly, the evaluation index of target detection method is designed. Through this paper, it is found that the success rate of some targets is naturally better than other targets, and easy to distinguish from the background. In order to improve the recognition success rate of the object in the complex environment and determine the position and attitude of the object, the pattern as the artificial identification in the environment is proposed. The interior art decoration pattern is selected as the experimental pattern and the pattern classification evaluation index is designed. The experimental results show that the method proposed in this paper can optimize the pattern subsets which are confused with each other and easy to distinguish from the background. It has a certain reference value for decorative pattern recognition in complex environment for COVID-19 epidemic.

Siamese Visual Tracking With Deep Features and Robust Feature Fusion

Trackers based on fully-convolutional Siamese networks regard tracking as a process of learning a similarity function. By utilizing shallow networks and off-line training, Siamese trackers can achieve high tracking speed and perform well in some simple scenes. However, due to the less semantic information and the invariant template, Siamese trackers still have a gap compared with the state-of-the-art methods in complex scenes and other challenging problems (Occlusion, Deformation, etc.). In this paper, we propose a Siamese tracking algorithm with deep features and robust feature fusion (SiamDF). The improved ResNet-18 network is utilized to replace the traditional shallow network and extract the deep features with more semantic information. For eliminating the negative effect of padding and making better use of the deep network, the proposed algorithm adopts the spatial aware sampling strategy to overcome the strict translation invariance. Meanwhile, a final response map with high quality can be obtained by using the multi-layer feature fusion. Thus, the tracker can significantly reduce the impact of the distractors in complex scenes. In addition, an adaptive feature information fusion is adopted to update the template, so that the algorithm can adapt to various changes of the target appearance. Objective evaluation on the OTB100 dataset shows that the precision and the overlap success can reach 0.852 and 0.658 respectively. Moreover, the EAO value evaluated on the VOT2016 database can reach 0.336. These results demonstrate that our algorithm can effectively improve the tracking performance and perform favorably in both robustness and accuracy.

Level set based shape prior and deep learning for image segmentation

Deep convolutional neural network can effectively extract hidden patterns in images and learn realistic image priors from the training set. And fully convolutional networks (FCNs) have achieved state-of-the-art performance in the image segmentation. However, these methods have the disadvantages of noise, boundary roughness and no prior shape. Therefore, this study proposes a level set with the deep prior method for the image segmentation based on the priors learned by FCNs. The FCNs can learn high-level semantic patterns from the training set. Also, the output of the FCNs represents the high-level semantic information as a probability map and the global affine transformation can obtain the optimal affine transformation of the intrinsic prior shape. Moreover, the improved level set method integrates the information of the original image, the probability map and the corrected prior shape to achieve the image segmentation. Compared with the traditional level set method of simple scenes, the proposed method solves the disadvantage of FCNs by using the high-level semantic information to segment images of complex scenes. Finally, Portrait data set are used to verify the effectiveness of the proposed method. The experimental results show that the proposed method can obtain more accurate segmentation results than the traditional FCNs.

A route navigation algorithm for pedestrian simulation based on grid potential field

Pedestrian simulation modeling has become an important means to study the dynamic characters of dense populations. In the continuous pedestrian simulation model for complex simulation scenario with obstacles, the pedestrian path planning algorithm is an indispensable component, which is used for the calculation of pedestrian macro path and microscopic movement desired direction. However, there is less efficiency and poor robustness in the existing pedestrian path planning algorithm. To address this issue, we propose a new pedestrian path planning algorithm to solve these problems in this article. In our algorithm, we have two steps to determine pedestrian movement path, that is, the discrete potential fields are first generated by the flood fill algorithm and then the pedestrian desired speeds are determined along the negative gradient direction in the discrete potential field. Combined with the social force model, the proposed algorithm is applied in a corridor, a simple scene, and a complex scene, respectively, to verify its effectiveness and efficiency. The results demonstrate that the proposed pedestrian path planning algorithm in this article can greatly improve the computational efficiency of the continuous pedestrian simulation model, strengthen the robustness of application in complex scenes.

An adaptive rendering framework for efficient synthetic light field generation

Real-time global illumination rendering is very desirable for emerging applications such as Virtual Reality (VR) and Augmented Reality (AR). However, client devices have difficulties to support photo-realistic rendering, such as Ray-Tracing, due to insufficient computing resources. Many modern frameworks adopted Light Field rendering to support device displaying. A Light Field can be precomputed and stored in the Cloud. During runtime, the display extracts the color information from the Light Field to generate arbitrary real time viewpoints or re-focusing within a predefined area. To efficiently compute the Light Field, We have combined DIBR (Depth-Image-Based-Rendering) and traditional ray-tracing in an adaptive fashion to synthesize images. By measuring the color errors during runtime, we adaptively determine a good balance between DIBR and Ray Tracing. We study several cases and problems when combing these algorithms and came up with a more efficient solution. To further optimize the computation efficiency, we also added a multi-level design to exploit the degree of shareable pixels among images to control the computation for error reduction. In addition, we also design an efficient multi-threading scheduling to show the scalability of our approach. Experiments show that when the number of threads is small, we achieved up to 3.24X speedup in Light Field generation for relative simple scenes like Cornell Box, and about 2X speed up for complex scenes like Conference Room or Sponza. In the multi-thread environment, our proposed scheme can offer about 2X speedup over traditional ones.

Application Research of Moving Target Detection Based on Optical Flow Algorithms

After many years of research, optical flow algorithm has achieved good results in detecting moving objects in simple scenes, but the detection effect in some complex scenes is not ideal, for example, in scenes with changing illumination and large displacement, the accuracy of moving objects detection is low. In order to solve this problem, this paper proposes texture decomposition of images, and applies texture image and pyramid technology to Lucas-Kanade optical flow algorithm. Relevant experiments show that this method can achieve better detection results for moving objects in static scenes.

Effective shadow removal via multi-scale image decomposition

Shadow removal is a fundamental and challenging problem in image processing field. Current approaches can only process shadows with simple scenes. For complex texture and illumination, the performance is less impressive. In this paper, we propose a novel shadow removal algorithm based on multi-scale image decomposition, which can recover the illumination for complex shadows with inconsistent illumination and different surface materials. Independent of shadow detection, our algorithm only requires a rough boundary distinguishing shadow regions from non-shadow regions. It first performs a multi-scale decomposition for the input image based on an illumination-sensitive smoothing process and then removes shadows in the basic layer using a local-to-global optimization strategy, which fuses all local shadow-free results in a global manner. Finally, we recover the texture details for the shadow-free basic layer and obtain the final shadow-free image. We validate the performance of the proposed method under various lighting and texture conditions and show consistent illumination between shadow and surrounding regions in the shadow removal results.

Trauma system resource preservation: A simple scene triage tool can reduce helicopter emergency medical services overutilization in a state trauma system.

Helicopter emergency medical services improve survival in some injured patients but current utilization leads to significant overtriage with considerable numbers of transported patients discharged home from the emergency department or found to have non-time-sensitive injuries. Current triage models for utilization are complex and untested. Data from a state trauma registry were reviewed from 1987 to 1993 and from 2013 to 2015 and compared. Data from 2013 to 2015 were analyzed for field information found to influence mortality and a model for low mortality-risk patients designed. Indexed to population, a major increase in numbers of injured patients transported directly to designated trauma centers (39.849-167.626/100,000/year) occurred with an increased portion transported by helicopter emergency medical services from 7.28% to 9.26%. A simple triage tool to predict low mortality rates was designed utilizing results from logistic regression. Nongeriatric adult patients (age, 16.0-69.9 years) with a blunt injury mechanism, normal Glasgow Coma Scale motor score, pulse rate of 60 bpm to 120 bpm and respiratory rate of 10 breaths per minute to 29 breaths per minute are at low risk for mortality. Cost for helicopter transportation was substantially higher than ground transportation based on available data. Cost differentials in transport mode increased patient financial risk when helicopter transportation was utilized. Implementing a simple decision tool designating nongeriatric adult patients with a blunt injury mechanism, normal Glasgow Coma Scale motor score, systolic blood pressure greater than 90 mm Hg, pulse rate of 60 bpm to 120 bpm, and respiratory rate of 10 breaths per minute to 29 breaths per minute to ground transportation would result in substantial savings without an increase in mortality and reduce risk of patient financial harm. Prognostic/Epidemiological study, level IV. Economic and value based evaluation, level IV.