We propose a linear regression model for the estimation of human body measurements. The input to only consists information that person can self-estimate, such as height and weight. evaluate our against state-of-the-art approaches measurement from point clouds images, demonstrate comparable performance with best methods, even outperform several deep learning models on public datasets. simplicity proposed makes it perfectly suitable baseline in addition convenience applications virtual try-on....

3D digital anthropometry is the study of estimating human body measurements from scans. Precise are important health indicators in medical industry, and guiding factors fashion, ergonomic entertainment industries. The measuring protocol consists scanning whole subject static A-pose, which maintained without breathing or movement during process. However, A-pose not easy to maintain process, can last even up a couple minutes. This constraint affects final quality scan, turn accuracy estimated...

The purpose of this study is to give a performance comparison between several classic hand-crafted and deep key-point detector descriptor methods. In particular, we consider the following classical algorithms: SIFT, SURF, ORB, FAST, BRISK, MSER, HARRIS, KAZE, AKAZE, AGAST, GFTT, FREAK, BRIEF RootSIFT, where subset all combinations paired into detector-descriptor pipelines. Additionally, analyze two recent perspective models, LF-Net SuperPoint. Our benchmark relies on HPSequences dataset that...

Single-shot dense 3D reconstruction using colored structured light is a difficult problem due to the undesired effects of ambient lighting, object albedo, non-equal channel gains, and cross-talk. We propose novel single-shot light. Our method combines self-equalizing De Bruijn sequence, scale-space analysis, bandpass complex Hilbert filters achieve insensitivity gains. The proposed reconstructs about 85% points compared time-multiplexing strategies decoding error in recovered projector...

3D scanning based on structured light (SL) has been proven to be a powerful tool measure the three-dimensional shape of surfaces, especially in biomechanics. We define set conditions that an optimal SL strategy should fulfill case static scenes and then we present efficient solution improving number-theoretic approach (NTA). The proposal is compared well-known Gray code (GC) plus phase shift (PS) technique original NTA, all satisfying same but obtaining significant improvements with our...

One of the very first tablets with a built-in DLP projector has recently appeared on market while smartphones have been available around for quite while. Interestingly, 3D reconstruction solutions mobile devices never considered exploiting implementation powerful active stereo concept, structured light (SL), whose main component is camera-projector pair. In this work we demonstrate framework implementing SL tablet. addition, propose registration method by taking advantage in novel way two...

Structured light 3D surface scanners are usually comprised of one projector and camera which provide a limited view the object's surface. Multiple projectors cameras must be used to reconstruct whole profile. Using multiple in structured profilometry is challenging problem due inter-projector interferences make pattern separation difficult. We propose use sinusoidal fringe patterns where each has its own specifically chosen set temporal phase shifts together comprise DFT2P+1 basis, P number...

Current 3D body scanners based on structured light principle are expensive and somewhat bulky machines which limits their wider spread.One of the factors driving scanner cost up is requirement to synchronize projector camera requires either system components a costly customized hardware synchronization solution.We propose software solution projector-camera problem enables construction low-cost using common commercial off-the-shelf only: projector, camera, computer.We also simple calibration...

Human body scanners using structured light are usually comprised of one projector which provides a limited view the surface due to field-of-view.To scan whole human multiple projectors must be used.Using more than in scanner is typically difficult inter-projector interferences make reconstruction hard task.We propose use temporal phase-shifts enable multiplexing fringe-projection profilometry.In our approach each projects sinusoidal fringe pattern having its own specifically chosen set phase...

In the case of structured data, such as 2D images, many variants traditional convolution neural network architectures have been successfully proposed. Learning from unstructured sets 3D point clouds, is a challenging task due to numerous reasons among which two most important ones are: cloud generally (i) unordered and (ii) sparse data set. Therefore, proposed are invariant both ordering number points in cloud. PointNet one architecture, originally introduced demonstrated on classification...

This study proposes a method to calibrate 3D kinematic systems. The authors briefly address typical wand-based calibration, that is, calibration using rigid bar, both from the computational point of view and procedural perspective, then they define their alternative way calibration. proposed takes advantage feature presently used tools, orthogonality wands. usual two steps in wand are reduced single one. In addition, propose an for how enforce typically available geometric constraint, known...

The use of one-shot pattern projection to obtain 3D dense reconstruction constitutes a promising field research in structured light. Most the related works presented literature are based on fringe extract depth from phase deviation. However, algorithms employed unwrap computationally slow and can fail under certain slopes occlusions object shape. In these lines, color using wavelet decomposition is presented. Moreover, novel unwrapping algorithm proposed, providing fast reliable absolute map...