Correcting the Eccentricity Error of Projected Spherical Objects in Perspective Cameras

Projective transformation of spheres onto images produce ellipses, whose centers do not coincide with the projected center of the sphere. This results in an eccentricity error, which must be treated in high precision metrology. This article provides closed formulations for modeling this error in images to enable 3-dimensional (3D) reconstruction of the center of spherical objects. The article also provides a new direct robust method for detecting spherical pattern in point clouds. It was shown that the eccentricity error in an image has only one component in the direction of the major axis of the ellipse. It was also revealed that the eccentricity is zero if and only if the center of the projected sphere lies on the camera’s perspective center. The effectiveness of the robust sphere detection and the eccentricity error modeling method was evaluated on simulated point clouds of spheres and real-world images, respectively. It was observed that the proposed robust sphere fitting method outperformed the popular M-estimator sample consensus in terms of radius and center estimation accuracy by a factor of 13, and 14 on average, respectively. Using the proposed eccentricity adjustment, the estimated 3D center of the sphere using modeled eccentricity was superior to the unmodeled case. It was also observed that the accuracy of the estimated 3D center using modeled eccentricity continuously improved as the number of images increased, whereas the unmodeled eccentricity did not show improvements after eight image views. The results of the investigation show that: (i) the proposed method effectively modeled the eccentricity error, and (ii) the effects of eliminating the eccentricity error in the 3D reconstruction become even more pronounced in a larger number of image views.

Deep-Learning-Based Non-Coherent DPSK Multiple-Symbol Differential Detection in Single-User Massive MIMO Systems

<div>In this work, non-coherent massive MIMO differential phase-shift keying modulation (DPSK) detection is considered to get rid of the complexity of channel estimation. However, most of the well-performing DPSK detection techniques require high computational complexity at the receiver. The use of deep-learning is proposed for detecting the transmitted DPSK symbols over a single-user massive MIMO system. We provide a multiple-symbol differential detection implementation using deep-learning. Two deep-learning-based multiple-symbol differential detection receiver designs are proposed and compared with differential detection (DD), decision-feedback differential detection (DFDD), and multiple-symbol differential detection (MSDD) for the same system parameters. Where multiple-symbol differential sphere detection (MSDSD) is used to implement MSDD. The results show that the proposed deep-learning-based classification neural networks outperform decision-feedback differential detection and achieve an optimal performance compared to conventional multiple-symbol differential detection implemented by multiple-symbol differential sphere detection.</div>

Deep-Learning-Based Non-Coherent DPSK Multiple-Symbol Differential Detection in Single-User Massive MIMO Systems

<div>In this work, non-coherent massive MIMO differential phase-shift keying modulation (DPSK) detection is considered to get rid of the complexity of channel estimation. However, most of the well-performing DPSK detection techniques require high computational complexity at the receiver. The use of deep-learning is proposed for detecting the transmitted DPSK symbols over a single-user massive MIMO system. We provide a multiple-symbol differential detection implementation using deep-learning. Two deep-learning-based multiple-symbol differential detection receiver designs are proposed and compared with differential detection (DD), decision-feedback differential detection (DFDD), and multiple-symbol differential detection (MSDD) for the same system parameters. Where multiple-symbol differential sphere detection (MSDSD) is used to implement MSDD. The results show that the proposed deep-learning-based classification neural networks outperform decision-feedback differential detection and achieve an optimal performance compared to conventional multiple-symbol differential detection implemented by multiple-symbol differential sphere detection.</div>

Model Penghindaran Tabrakan Multi Obyek Menggunakan Repulsive Field

Mendeteksi tabrakan antar obyek dan menampilkan gerakan penghindaran tabrakan yang alami menjadi topik yang diinginkan dalam berbagai aplikasi navigasi agen otonom. Pada umumnya, gerakan penghindaran tabrakan multi obyek statis dan dinamis secara alami banyak menggunakan teknik dengan pengambilan keputusan untuk bergerak menghindar kearah lain. Teknik penghindaran ini mengalami kesulitan ketika jumlah obyek diperbanyak. Pada penelitian ini menyajikan teknik yang mampu menampilkan gerakan penghindaran tabrakan multi obyek statis dan dinamis secara alami. Metode yang digunakan Sphere-Plane Detection (SPD) dan Sphere-Sphere Detection (SSD) untuk mendeteksi tabrakan dengan referensi jarak dan metode potential field jenis repulsive untuk penghindaran tabrakan. Pengujian dilakukan dengan jumlah obyek 100 yang terdiri dari 34 agen A, 33 agen B dan 33 agen C, multi penghalang dengan dan tanpa repulsive, penghalang statis dan dinamis, serta diuji dalam animasi boid. Dari pengujian yang telah dilakukan diperoleh nilai vektor |V| dari posisi awal sampai ke tujuan untuk agen A2 sebesar 51.4522, agen B1 bernilai 45.0853 dan agen C4 sebesar 27.7237, setiap agen memiliki lebih dari satu jalur untuk menuju tujuan, didapatkan tiga model gerakan penghindaran dan tiga parameter yang mempengaruhi gerakan agen yaitu perubahan nilai repulsive, jarak antar penghalang dinamis.

Monitoring of the Production Process of Graded Concrete Component Using Terrestrial Laser Scanning

Accepting the ecological necessity of a drastic reduction of resource consumption and greenhouse gas emissions in the building industry, the Institute for Lightweight Structures and Conceptual Design (ILEK) at the University of Stuttgart is developing graded concrete components with integrated concrete hollow spheres. These components weigh a fraction of usual conventional components while exhibiting the same performance. Throughout the production process of a component, the positions of the hollow spheres and the level of the fresh concrete have to be monitored with high accuracy and in close to real-time, so that the quality and structural performance of the component can be guaranteed. In this contribution, effective solutions of multiple sphere detection and concrete surface modeling based on the technology of terrestrial laser scanning (TLS) during the casting process are proposed and realized by the Institute of Engineering Geodesy (IIGS). A complete monitoring concept is presented to acquire the point cloud data fast and with high-quality. The data processing method for multiple sphere segmentation based on the efficient combination of region growing and random sample consensus (RANSAC) exhibits great performance on computational efficiency and robustness. The feasibility and reliability of the proposed methods are verified and evaluated by an experiment monitoring the production of an exemplary graded concrete component. Some suggestions to improve the monitoring performance and relevant future work are given as well.

IMPLEMENTASI MIMO DECODER BERBASIS METODE NON-LINEAR UNTUK MENINGKATKAN KINERJA WLAN 802.11ac

Kebutuhan sistem komunikasi nirkabel yang mendukung laju data dan kinerja yang semakin tinggi mendorong perkembangan teknologi WLAN. Teknologi ini mengerucut pada penggunaan teknologi terkini, yaitu MIMO-OFDM. Implementasi OFDM pada perkembangan WLAN dimulai sejak IEEE 802.11a, hingga saat ini mencapai 900 Mbps pada IEEE 802.11ac. Dengan penggunaan kanal yang sama untuk beberapa data pada setiap antenna, dibutuhkan teknik khusus untuk mendapatkan kembali informasi yang dikirim. Dua teknik yang umum digunakan adalah berbasis metode linear, yaitu ZF dan MMSE. Keduanya memiliki kompleksitas yang rendah, tetapi kinerjanya juga rendah. Teknik yang dikenal optimal adalah berbasis non-linear, yaitu MLD. Teknik ini memiliki kinerja paling baik tetapi tingkat kompleksitasnya tinggi. Untuk menengahi kedua metode tersebut, dikembangkanlah metode non-linear yang sub-optimal, seperti K-Best, Trellis, dan Sphere Detection. Ketiga metode ini memiliki kompleksitas yang rendah dan menghasilkan kinerja yang baik. Pada penelitian ini akan dilakukan simulasi dan analisa terhadap kinerja MIMO Decoder yang berbasis metode non-linear, yaitu K-Best, Trellis, dan Sphere Detection. Simulasi dilakukan pada MCS 5, 6, dan 7 dengan konfigurasi 6x6 dan 40 MHz bandwidth. Konfigurasi ini akan menghasilkan laju data sebesar 720 Mbps, 810 Mbps, dan 900 Mbps.