RGB-D Sensors Data Quality Assessment and Improvement for Advanced Applications

Rodríguez‐Gonzálvez, Pablo; Guidi, Gabriele

doi:10.1007/978-3-030-28603-3_4

Cited by 4 publications

(5 citation statements)

References 58 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These results are also compared with the performances of other close-range sensors in the market. In the book RGB-D Image Analysis and Processing [ 32 ], Rodríguez and Guidi [ 33 ] dedicate one chapter to data quality assessment and improvement for RGB-D cameras, analyzing how random and systematic 3D measurement errors affect the global 3D data quality in the various operating principles. An experimental evaluation of three state-of-the-art Intel ® RealSense™ cameras was performed by Lourenço and Araujo [ 34 ].…”

Section: Related Workmentioning

confidence: 99%

“…The differential in AS lies in the features added to the scene through an unstructured light pattern. These artificial features are significant for low-texture surfaces, allowing a better reconstruction once we have many more correspondence points to triangulate [ 30 , 33 ]. The triangulation principle applies in a similar manner as in SL, so the Equation ( 1 ) also applies to AS.…”

Section: Fundamentalsmentioning

confidence: 99%

See 1 more Smart Citation

An Experimental Assessment of Depth Estimation in Transparent and Translucent Scenes for Intel RealSense D415, SR305 and L515

Curto

Araújo

2022

Sensors

View full text Add to dashboard Cite

RGB-D cameras have become common in many research fields since these inexpensive devices provide dense 3D information from the observed scene. Over the past few years, the RealSense™ range from Intel® has introduced new, cost-effective RGB-D sensors with different technologies, more sophisticated in both hardware and software. Models D415, SR305, and L515 are examples of successful cameras launched by Intel® RealSense™ between 2018 and 2020. These three cameras are different since they have distinct operating principles. Then, their behavior concerning depth estimation while in the presence of many error sources will also be specific. For instance, semi-transparent and scattering media are expected error sources for an RGB-D sensor. The main new contribution of this paper is a full evaluation and comparison between the three Intel RealSense cameras in scenarios with transparency and translucency. We propose an experimental setup involving an aquarium and liquids. The evaluation, based on repeatability/precision and statistical distribution of the acquired depth, allows us to compare the three cameras and conclude that Intel RealSense D415 has overall the best behavior namely in what concerns the statistical variability (also known as precision or repeatability) and also in what concerns valid measurements.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Fundamentalsmentioning

confidence: 99%

An Experimental Assessment of Depth Estimation in Transparent and Translucent Scenes for Intel RealSense D415, SR305 and L515

Curto

Araújo

2022

Sensors

View full text Add to dashboard Cite

show abstract

“…The stability and noise characteristics of the raw data were evaluated by computing the Allan deviation of the phases calculated according to (3). This computation was based on a time-series of 8000 samples acquired over two hours on a static scene composed of a flat white plane at a distance of about 1 m using the 1.5 m operating mode.…”

Section: Performance Evaluationmentioning

confidence: 99%

“…Recent rapid growth of interest in robotically assisted construction [1] has also boosted the use of sensing technologies to acquire processrelevant information. These works showcased the potential of various sensors in digital fabrication processes and demonstrated ways in which accurately measured 3D information and object parameters extracted from such information can be used for in-line process improvement via feedback control [2,3]. In order to extract the relevant information for construction processes, appropriate processing algorithms and interpretation of the 3D data need to be employed.…”

Section: Introductionmentioning

confidence: 99%

Depth-Camera-Based In-line Evaluation of Surface Geometry and Material Classification For Robotic Spraying

Frangež¹,

Salido-MonzÃo²

2020

Proceedings of the 37th International Symposium on Automation and Robotics in Construction (ISARC)

View full text Add to dashboard Cite

This paper presents a feasibility study of surface geometry (SG) evaluation and material classification (MC) for robotic spraying. We propose two complementary approaches using point clouds and intensity data provided by a state-of-the-art industrial time-of-flight (ToF) depth camera. The SG evaluation is based on geometric feature computation within local neighbourhoods, which are then used within a supervised classification. The results of this approach are SG classes according to the level of geometric variability of the surface, displayed as SG maps. For MC, active reflectance estimation is investigated and exploited to derive features related to the reflectance and diffusive properties of each material for classification. The result of both approaches can be prospectively used as feedback in digital fabrication for in-line adaptation of the process to improve control of relevant geometrical and material properties.

show abstract

“…The latter starts to get attention in published literature, where fabrication setups augmented with various sensing systems show potential of in-line process improvement via feedback control, as e.g. demonstrated in Wolfs (2019), Rodríguez-Gonzálvez and Guidi (2019), Bard et al (2018). In particular, this feedback is relevant in additive manufacturing applications with cementitious materials which are unpredictable and hard to model in their dynamic transition from the soft to hard state.…”

Section: Introductionmentioning

confidence: 99%

Geometric Feedback System for Robotic Spraying

Frangež¹,

Taha²,

Feihl³

et al. 2022

PFG

View full text Add to dashboard Cite

In this paper, we tackle the task of replacing labor intensive and repetitive manual inspection of sprayed concrete elements with a sensor-based and automated alternative. We present a geometric feedback system that is integrated within a robotic setup and includes a set of depth cameras used for acquiring data on sprayed concrete structures, during and after fabrication. The acquired data are analyzed in terms of thickness and surface quality, with both sets of information then used within the adaptive fabrication process. The thickness evaluation is based on the comparison of the as-built state to a previous as-built state or to the design model. The surface quality evaluation is based on the local analysis of 3D geometric and intensity features. These features are used by a random forest classifier trained using data manually labelled by a skilled professional. With this approach, we are able to achieve a prediction accuracy of 87 % or better when distinguishing different surface quality types on flat specimens, and 75 % when applied in a full production setting with wet and non-planar surfaces. The presented approach is a contribution towards in-line material thickness and surface quality inspection within digital fabrication.

show abstract

RGB-D Sensors Data Quality Assessment and Improvement for Advanced Applications

Cited by 4 publications

References 58 publications

An Experimental Assessment of Depth Estimation in Transparent and Translucent Scenes for Intel RealSense D415, SR305 and L515

An Experimental Assessment of Depth Estimation in Transparent and Translucent Scenes for Intel RealSense D415, SR305 and L515

Depth-Camera-Based In-line Evaluation of Surface Geometry and Material Classification For Robotic Spraying

Geometric Feedback System for Robotic Spraying

Contact Info

Product

Resources

About