Urban Street Scenes Research Articles

Super-Segments Based Classification of 3D Urban Street Scenes

We address the problem of classifying 3D point clouds: given 3D urban street scenes gathered by a lidar sensor, we wish to assign a class label to every point. This work is a key step toward realizing applications in robots and cars, for example. In this paper, we present a novel approach to the classification of 3D urban scenes based on super-segments, which are generated from point clouds by two stages of segmentation: a clustering stage and a grouping stage. Then, six effective normal and dimension features that vary with object class are extracted at the super-segment level for training some general classifiers. We evaluate our method both quantitatively and qualitatively using the challenging Velodyne lidar data set. The results show that by only using normal and dimension features we can achieve better recognition than can be achieved with high-dimensional shape descriptors. We also evaluate the adopting of the MRF framework in our approach, but the experimental results indicate that thisbarely improved the accuracy of the classified results due to the sparse property of the super-segments.

Visual evaluation of urban streetscapes: How do public preferences reconcile with those held by experts?

Visual qualities of well-liked settings are known to include richness and variety, tempered by perceptible underlying structure and clear associational meanings. For a variety of reasons – among them, technological, economic and professional principles – the design coherence of the built environment appears to have diminished in the wake of Modernism. However, opinions about the quality of the built environment are usually voiced by experts. Does it follow that their opinions are also those of lay people? Indeed, looking beyond academic critique of the appearance of the built environment, what are the views and aesthetic preferences of those who use the city? The article reports on the results of recent research that has evaluated perceptions of and preferences for a range of different urban street scenes. The opinions of the lay public as well as design and planning professionals have been collected for comparison. The research seeks to understand whether there are built-form characteristics, seen across a collection of buildings that make up a street edge, that are preferred over others. Based on data from 200 survey responses and a focus group discussion, the findings identify characteristics that are deemed to foster well-liked urban settings as well as those that should be avoided.

Cultural animation and economic vitality: Identifying the links and regeneration potential through the lens of the urban street scene

Culture, space and economy are intermeshed in complex ways. This paper reports on findings from a larger empirical research project commissioned to investigate the symbiotic relationship between culturally animated urban street scenes and economic vitality. Grounded in empirical qualitative research focusing on recent place quality enhancement schemes in the North East of England, the central aim of this paper is to make the case that everyday cultural activity and economically vibrant places can go hand in hand. The research did not seek to quantify the economic benefits of investments in the cultural animation of urban space, but interpretive analysis suggests that place quality regeneration strategies can enhance the economic performance and vitality of places. Based on the argument that cultural production of space and economic development are not, and therefore should not be viewed as, competing objectives, the paper puts forward a range of good practice pointers for policy makers and practitioners embarking on place quality enhancement schemes.

Sargent and Italy

This extravagantly illustrated catalogue - published in association with a major exhibition - evokes the romantic fascination with Italy that glimmers in the work of John Singer Sargent. Sargent, heralded on both sides of the Atlantic, was one of the most creative American artists of the late nineteenth century. Born in Florence to American parents living abroad, he retained a deep and lifelong connection to the country famed for its ability to get 'ineradicably in one's blood'. Sargent vacationed frequently in Italy, and most of the works he created there were painted not for commission but out of his artistic passion for Italy's people, land, and culture. Often hauntingly powerful, they range from dramatically painted genre scenes of Italian peasants and saturated landscapes that celebrate the beauty of the Italian countryside to portraits of other Anglo-American expatriates and tourists, including Henry James and Edith Wharton.The majority of works are of Italian sites, including well-known tourist spots but also the quieter, more isolated locales that Sargent sought out. His subjects include magnificent Italian gardens with their ancient and Baroque statuary, Rome's Neoclassical and Renaissance buildings, urban street scenes, the Italian Alps, and, of course, Venetian canals. Sargent found Venice particularly alluring, and the city well suited the watercolor medium in which he worked most often in Italy. His use of vivid colors, brushwork that varied from soft and fluid to bold and dashing, and an overwhelming sense of light and air characterize his Italian scenes - and rank Sargent as one of the finest watercolorists of all time. His later Italian works, some in watercolor and others in oil, reveal an artist who relished his materials and made art purely for art's sake.Both beautiful and informative, this lavish volume includes eighty-five color and fifty black-and-white images. It adds a new dimension to our appreciation of Sargent's art and will delight anyone who loves Italy, as Sargent so passionately did.

Stereo- and neural network-based pedestrian detection

Pedestrian detection is essential to avoid dangerous traffic situations. We present a fast and robust algorithm for detecting pedestrians in a cluttered scene from a pair of moving cameras. This is achieved through stereo-based segmentation and neural network-based recognition. The algorithm includes three steps. First, we segment the image into sub-image object candidates using disparities discontinuity. Second, we merge and split the sub-image object candidates into sub-images that satisfy pedestrian size and shape constraints. Third, we use intensity gradients of the candidate sub-images as input to a trained neural network for pedestrian recognition. The experiments on a large number of urban street scenes demonstrate that the proposed algorithm: (1) can detect pedestrians in various poses, shapes, sizes, clothing, and occlusion status; (2) runs in real-time; and (3) is robust to illumination and background changes.

Feature vectors for road vehicle scene classification

Effective feature vectors are needed if artificial neural networks (ANNs) are to provide an efficient method for classifying an area of interest (AOI) in a visual image captured from closed circuit television systems (CCTV) with the camera static. Problems of object orientation (rotation) and scale (size variation due to viewing distances) in an image are important in moving object classification or recognition. In this paper, feature vectors for the subimage moving objects associated with urban road scenes have been derived from invariant geometric properties of the object and statistical parameters of the image scene. These have been used as inputs for a neural network classifier. The extension of binary moment invariant techniques to grey level moment invariants has provided enhanced feature vectors for classification of objects in such images and shown to be effective. Anomalies such as moving shadows, swaying trees, changing reflectivity, etc., could be overcome by suitable preclassification neural stages. It has been shown that incorporating these anomalies along with the objects, resulting in a one-stage system, can give the best result, particularly when the training data for anomalies is limited. The results from the actual application of these techniques to classification of the moving objects of a road scene into groups and remaining anomalies is presented and analysed.

Road recognition with a neural network

This paper presents a neural-network model that is capable of recognising roads in colour images of urban and rural road scenes. The network is trained to classify the visible objects in an image, given a set of hand-labelled examples. The two problems of segmentation and recognition are separated through the use of “ideal” segmentations, allowing the performance of the recognition method to be assessed independently of the effects of using an imperfect real segmentation process. The generalisation performance of the method is measured using unseen test data. Results for both on-road and offroad viewpoints are presented. The system is found to recognise correctly approx. 70% of the regions, and approx. 90% of the area of the images

Road Recognition with A Neural Network

This paper presents a neural network model that is capable of recognising roads in colour images of urban and rural road scenes. The network is trained to classify the visible objects in an image given a set of hand-labelled examples. The two problems of segmentation and recognition are separated through the use of “ideal” segmentations, allowing the performance of the recognition method to be assessed independently of the effects of using an imperfect real segmentation process. Results for both on-road and off-road viewpoints are presented

UTD propagation model in an urban street scene for microcellular communications

A three-dimensional propagation model for microcellular communications in an urban street scene is presented. The model is based on the uniform theory of diffraction (UTD) and takes into account multiple wall-to-wall, wall-to-ground, and ground-to-wall reflections, the diffraction from corners of buildings, and subsequent reflections from such diffracted signals. The ray geometry is made extremely complex by the presence of ground reflections and the many combinations of sequences of reflections or diffractions from walls, edges, and ground. At each reflection or diffraction point, the local ray-fixed coordinate system or edge-fixed coordinate system is used together with appropriate dyadic reflection or diffraction coefficient matrices. The theoretical results for the signal path loss along the streets are compared with measurements done in New York and Tokyo for various values of the propagation parameters. Agreement with these measurements indicates that the UTD formulation is a good model for such urban communication applications. >

Visual Vulnerability of Streetscapes to Elevated Structures

Overhead expressways have damaged streetscapes in some countries, and new transportation system structures will possibly repeat the same failure. To prevent this, transportation planners must know which streetscapes are vulnerable to elevated structures and which are resistant to them. Hence, the relationship between streetscape type and visual vulnerability to an overhead guideway and an elevated station of new transportation systems was studied experimentally. Images of elevated structures were superimposed upon selected photographs of urban and suburban street scenes and presented to respondents. Using respondents' rating scores, the separate effects of guideway, station, and background scene were analyzed. We found that the major determinant of visual vulnerability is not visual quality of streetscapes but visual type; therefore, it is possible to judge visual vulnerability of streetscapes from visual type.

Conspicuity of traffic control devices : Cole, B.L., and Jenkins, S.E.Australian Road Research, 1982, 12.4, 223–238

This paper is concerned with obtaining: estimates of the conspicuity of traffic control devices seen in ordinary road environments, and an indication of the target object properties that determine conspicuity. The principal measure of conspicuity chosen was the probability of seeing a target object under defined conditions of fixation, observation time, and a second measure was in terms of the order in which the objects were first noticed. Experiments were conducted using projected color slides of 100 urban and semi-urban road scenes. It was found that traffic control devices were not conspicuous in the above terms. Size was an important determinant of conspicuity. Edge definition was also important. Brightness was less effective and white signs were noticed less frequently than colored ones. Language: en

Visual Preferences in Urban Street Scenes

This research examined visual environmental preferences of a sample of 46 subjects (Ss): 29 in Japan and 17 in the United States. The Ss (who were all students) were asked to evaluate videotapes and slides of urban street scenes in each country. An examination of preference scores revealed that both Japanese and U.S. Ss preferred foreign scenes to native ones. Physical properties of each scene were judged by four professionals in the United States. An examination of correlations between preferences for scenes and physical properties of those scenes revealed that for both Japanese and U.S. Ss, preference was associated independently with ratings of upkeep, prominence of nature, and absence of vehicles. Stepwise regressions revealed preference to be a function of two factors: order (order, naturalness, and upkeep) and diversity (high contrast, diversity, and few vehicles).