Earticle

In mold structure machining, tool interference is a phenomenon which results from a collision between a blade of tool and a workpiece. Also tool collision is a phenomenon which results from a collision of holder with the object to be machined. These phenomena not only cause damages to mold and tool but also increase machining time and cost. To detect a collision of a tool to mold structure, first of all, the mold structure and a tool must be defined with famous geometric models such CSG, B-rep, and Voxel. A tool is defined as a combination of the blade, the shank, and the holder. This thesis reviews various collision detection algorithms using z-map and computer 3D graphic collision detection algorithms for the tool in machining a mold structure.

차량 내 수 많은 센서 노드들이 무선 데이터를 전송할 때 충돌을 검출하고 해결하기 위한 매체접속제어 프로토콜을 제안한다. 제안 프로토콜은 데이터 전송 시작 후 랜덤하게 충돌 검출 슬롯을 하나 선택하여 전송을 잠시 멈추고, 해당 채널을 센싱하여 다른 노드와의 충돌 여부를 에너지 레벨과 jam 신호 검출을 통하여 확인한다. 제안 방안은 이러한 충돌 검출 과정을 여러 번 수행함으로써 각 충돌 검출 단계에서 충돌을 인지한 노드는 걸러지고, 다음 단계에서는 살아남은 노드만 경쟁을 계속하여 충돌 해결 확률을 획기적으로 증가시킨다. 시뮬레이션 결과, 다수의 충돌 검출 단계를 사용한 제안 방식이 기존 경쟁 프로토콜보다 우수한 처리율을 갖는다. 또한 충돌 검출 단계의 개수와 충돌 검출 슬롯의 개수에 따라 제안 방안의 성능 변화를 살펴보고 최적 파라미터를 도출하였다.

This paper proposes a medium access control protocol for collision detection and resolution when a large number of sensor nodes transmits data in vehicle. The proposed protocol selects a random collision detection (CD) slot after data transmission, suspends its transmission and senses the channel to check whether a collision occurs by the detection of both energy level and jam signal. The proposed scheme uses multiple CD phases and in each CD phase, colliding stations are filtered and only surviving stations compete again in the next CD phase; thus, the collision resolution probability significantly increases. Simulation results show that the proposed protocol using the multiple CD phases has significantly better throughput than the conventional protocol. In addition, according to the number of CD phases and the number of CD slots per phase, the throughput aspect of the proposed scheme is investigated and the optimal parameters are derived.

본 논문에서는 가상 공간에서 객체와의 상호 작용을 위한 빠른 충돌 감지 방법을 제안하였다. 먼저 메쉬 그룹화 단계에서는, 전체 공간을 일정한 크기의 소공간들로 분할하고, 각각의 소공간에 속하는 메쉬들에 대하여 이들을 모두 포함하는 최소의 기본 입체도형(육면체 혹은 구)의 위치와 크기를 정함으로써 그룹화한다. 충돌 검출 단계에서는, HIP(Haptic Interface Point)가 어떤 그룹을 대표하는 입체도형 내부에 들어 있는지를 검사하여 특정 입체도형과의 충돌이 확인되면 해당 그룹의 메쉬들만을 대상으로 실제 충돌이 일어난 메쉬를 찾는다. 본 논문에서 제안한 방법을 적용한 경우와 적용하지 않은 경우의 연산시간을 측정 및 비교함으로써 제안한 알고리즘의 효용성을 확인하였다.

In this paper, we propose a fast collision detection method for interacting with objects in virtual space. First, in the mesh grouping step, the entire space is divided into small spaces of a predetermined size, and the positions and sizes of the smallest basic stereoscopic shapes (cube or sphere) including all of the meshes belonging to each small space are determined. In the collision detection step, it is checked whether a haptic interface point (HIP) is included in a three-dimensional figure representing a group. When a collision with a specific three-dimensional figure is confirmed, searching is performed only for the meshes in the group to find a mesh on which a possible real collision with HIP occurred. The effectiveness of the proposed algorithm is verified by measuring and comparing the computation time of the proposed method with and without the proposed method.

GPU는 메모리 대역폭이 연산 속도를 결정하는 병목 지점이 된다. 즉, GPU 프로그래밍 시에는 불규칙적인 메모리 액세스나 다중 스레드들 사이에서의 서로 다른 명령 실행 분기가 발생하면 속도가 크게 저하되는 문제가 발생한다. 따라서 게임 엔진 충돌 처리용으로 사용되는 kd-tree와 같은 적응형 탐색(adaptive traverse) 기법은, 불규칙적인 메모리 액세스 및 서로 다른 명령 분기로 인해 지금까지 GPU 구조에 적합하지 않은 것으로 인식되어 왔다. 그러나 최근 NVIDIA의 Fermi 아키텍처의 등장과 함께 CPU에서처럼 GPU 다중 프로세서에도 캐시 메모리가 적용되고 있다. 본 논문에서는 이러한 새로운 GPU 아키텍처의 장점을 활용해서 충돌 처리 시간을 크게 줄일 수 있는 GPU 기반 kd-tree를 제안한다. 제안하는 GPU 기반 병렬 kd-tree는 체크 지점 65536 개에서 최근접 삼각형까지의 거리를 찾는 작업이 Fermi 아키텍처(캐시 적용) 기반에서 단일 코어 CPU 기반 kd-tree에 비해 평균 백 만 배 이상(1.0x106) 빨라졌으며, 이전 세대 Tesla 아키텍처(캐시 미적용) 기반 병렬 kd-tree에 비해서도 약 50 배 가까이 빠른 속도를 보였다.

GPU has a performance bottleneck in memory bandwidth; GPU performance becomes degenerated when irregular memory access patterns and different branching occur. In this regard, it had been known that adaptive traverse methods including kd-tree, which inevitably cause irregular memory access and branching patterns were not suitable for GPU programming. However, with recent advance in GPU hardware architecture, NVIDIA’s Fermi architecture in particular, it is possible to embed cache memories for multi-cores in GPU. In this paper, we present a parallel kd-tree based on GPU for fast collision detection. Our method reduces time in detecting collisions for around one million fold for test meshes than kd-tree implemented in a single CPU core. Besides, the kd-tree implemented in newer Fermi architecture is around 50 times faster than the previous generation Tesla architecture thanks to multiprocessor memory caches.

동적인 산업 현장에서 작업자의 안전을 보장하기 위해서는 신속하고 정확하며 비용 효율적인 충돌 감지 시스템이 필수적이다. LiDAR나 RGB-D 센서를 기반으로 하는 기존의 솔루션들은 높은 정밀도를 제공하지만, 비용이 많이 들고 유지보수가 까다로우며 조명 불량, 가림, 반사 표면과 같은 환경 조건에 민감하다는 단점이 있다. 이러한 한계를 극복하기 위해, 본 연구에서는 단일 RGB 카메라만을 사용하는 경량 충돌 감지 프레임워크를 제안한다. 이 시스템은 객체 탐지를 위한 YOLOv8, 인체 자세 추정을 위한 ViTPose, 그리고 단안 깊이 추정을 위한 Depth Anything을 통합하여 인체 관절과 객체 중심의 3D 위치를 실시간으로 재구성한다. 전용 충돌 감지 알고리즘은 이러한 3D 좌표 간의 유클리드 거리를 계산하며, 보간 기법 및 관절 구조 기반의 이상치 보정을 통해 시스템의 안정성과 신뢰성을 향상시킨다. 제안된 프레임워크는 다양한 조명 조건, 카메라 각도 및 상호작용 유형을 포함한 55개의 시나리오 비디오로 구성된 자체 데이터셋을 통해 검증되었으며, 실험 결과 모든 시나리오에서 F1-score 1.0을 달성하여 높은 정확도와 강건성을 입증하였다. 이러한 연구 결과는 단일 RGB 카메라 시스템이 고가의 센서 기반 접근 방식을 대체할 수 있는 실용적이고 확장 가능하며 비용 효율적인 대안임을 보여주며, 이를 통해 산업 안전 모니터링을 위한 실시간 충돌 감지가 가능함을 시사한다.

Ensuring worker safety in dynamic industrial environments requires collision detection systems that are fast, accurate, and cost-efficient. Conventional solutions based on LiDAR or RGB-D sensors deliver high precision but are costly, maintenance-heavy, and sensitive to environmental conditions such as poor lighting, occlusion, and reflective surfaces. To address these limitations, this study proposes a lightweight collision detection framework using only a single RGB camera. The system integrates YOLOv8 for object detection, ViTPose for human pose estimation, and Depth Anything for monocular depth estimation to reconstruct 3D positions of human joints and object centers in real time. A dedicated collision detection algorithm computes Euclidean distances between these 3D coordinates, while temporal smoothing and joint-structure-based outlier correction enhance stability and reliability. The proposed framework was validated on a custom dataset of 55 scenario videos under diverse lighting conditions, camera angles, and interaction types. Experimental results show that the system achieved an F1-score of 1.0 across all scenarios, indicating high accuracy and robustness. These findings demonstrate the potential of single RGB camera systems as practical, scalable, and cost-effective alternatives to expensive sensor-based approaches, enabling real-time collision detection for industrial safety monitoring.

본 논문에서는 3D 그래픽에서 빠르고 정확한 충돌검사(Collision-detection)는 공간에서 표준객체를 중심으로 하는 연구가 많이 이루어져 왔다. H/W를 중심으로 3D그래픽 분야의 놀라운 발달과 다양한 충돌 알고리즘 연구를 통하여 3D 객체의 충돌 속도의 향상뿐만 아니라, 최근에는 사실적인 표현에 깊은 관심을 가지고 있다. 이 논문에서는 기존의 연구되어진 표준 3D 객체의 충돌 알고리즘을 특징을 분석하고, 단순한 계층 구조에서 LOD(Level-of-Detail)를 적용한 유향상자에 가변적 밀도의 알고리즘을 제안한다. 이 알고리즘을 이용하여 공간상에서 LOD 알고리즘을 기반으로 하여, 단순히 거리에 따라서 상세화와 간략화를 적용하는 방법에서 탈피하여 유향상자의 밀도의 알고리즘을 차별적으로 적용하여, 3D객체가 공간상에서 충돌검사의 성능을 향상 시켜, 3D 게임에서 효율적인 렌더링과 사실적인 표현을 제안하여 많은 3D 객체의 실시간을 중요시 하는 3D 게임에서 효율성을 높인다.

The advancement in hardware and research in various algorithms, the expressions of 3D objects that were once based on frames have proceed into high polygon rendering techniques, allowing increasingly realistic expressions. However, there still exist limitations in applying the high polygon process to real-time 3D games. This paper is intended to improve object collision detection performance by propose a simplification of the object in Bounding Volume(BV), which is commonly used in 3D games for collision detection among objects, through a "Level of Detail(LOD)" at each LOD step, and a test criteria based on "Flex-Density(FD)." This proposal can be utilized for 3D games that improvement in speeds and in realism, This algorithm can be further applied to various fields of 3D graphics that require real-time processing.

본 논문에서는 마커 기반의 가상 객체나 opengl 을 이용하여 구현된 가상 객체, 웹 카메라 영상 프레임을 통해 입력된 실물 객체의 충돌처리를 유효화시켜 사용자에게 보다 높은 현실성을 제공하는 증강현실 충돌기법을 제안한다. 증강현실 시스템은 현실의 환경에 가상의 객체를 추가하여 사용자에게 시각적 정보를 제공하는 시스템으로 객체간의 상호작용, 사용자와의 상호작용에 대하여 관심이 높아가고 있다. 그러나 본 논문에서 제안한 충돌처리는 상호작용에 있어서 가장 기본적인 문제가 되기 때문에 선결해야 할 문제이다. 따라서 제안된 시스템에서는 이 문제를 해결하여 마커를 기반으로 구현된 객체와 Opengl을 기반으로 구현된 가상 객체, 웹 카메라 영상 프레임으로부터 입렵된 실물 객체 사이에서 객체간의 충돌을 유효화 시킴으로써 현실성과 몰입도를 증대시킨다.

This paper suggests a collision techniques for a higher reality in augmented reality by processing collision between a real object obtained through video frame input and a marker-based virtual object or a virtual object from Opengl. Augmented reality system is providing the visual information containing a virtual object added to the real environment and interactive interface between objects and between user and objects becomes a more concerning interest. But the collision problem is essential to the interactive interface and has to be solved first. Therefore, the proposed system suggests a solution for it to increase the realism and the immersion by validating the collision among a marker-based object, a virtual object from Opengl, and a real object obtained through web camera, that is, video frame.

Smart Door(SD) is a human friendly power-assisted door system for passenger car doors. It offers comfort and safety to passengers and drivers by supplying additional power. In this study, dynamic system model and the equation of motion derivation are derived. And we propose the disturbance observer based collision detection algorithm for safety when opening the door. A disturbance caused by collision has a fast response compared to a friction, uncertainties and so on. The main idea this study is to estimate a variation of disturbance for stably and effectively detecting a collision. In order to evaluate a performance of collision detection, an experiment set up is constructed. The experimental results validate the usefulness of the proposed collision detection algorithm.

In case of South Korea, the airspace which airlines can operate is extremely limited due to the military operational area located within the Incheon flight information region. As a result, safety problems such as mid-air collision between aircraft or Traffic alert and Collision Avoidance System Resolution Advisory (TCAS RA) may occur with higher probability than in wider airspace. In order to prevent such safety problems, an mid-air collision risk detection model based on Detect-And-Avoid (DAA) well clear metrics is investigated. The model calculates the risk of mid-air collision between aircraft using aircraft trajectory data. In this paper, the practical use of DAA well clear metrics based model has been validated. Aviation safety data such as aviation safety mandatory report and Automatic Dependent Surveillance Broadcast is used to measure the performance of the model. The attributes of individual aircraft track data is analyzed to correct the threshold of each parameter of the model.

본 논문에서는 3D 그래픽에서 빠르고 정확한 충돌검사(collision-detection)는 3D공간에서 표준객체를 중심으로 하는 연구가 많이 이루어져 왔다. 3D그래픽 분야에서 H/W의 놀라운 발달과 다양한 3D그래픽 관련 논문에서 3D객체의 충돌 속도의 성능 향상뿐만 아니라 사실적인 표현에 깊은 관심을 가지고 있다. 3D 그래픽 알고리즘 중에서 표준 3D 객체의 다양한 충돌 알고리즘을 특징을 분석하고, 기존의 3D 객체의 단순한 계층 구조에서 LOD(Level-of-Detail)를 이용한 알고리즘를 제안한다. 이 알고리즘을 이용하여 3D공간상에서 LOD(Level-of-Detail) 알고리즘을 적용시켜서, LOD단계가 높은 (가까운) 곳에서는 객체의 유향상자를 자세히 검사하고, LOD단계가 낮은(먼곳)에 위치한 객체의 유향상자는 간략히 검사를 적용3D객체가 3D 공간상에서 충돌검사의 성능을 향상시키고 3D 그래픽에서 중요한 요소인 3차원 공간상의 효율적인 렌더링과 사실적인 표현을 제안하여 실시간을 중요시 하는 3D 게임에서 사실감과 효율성을 높였다.

In this paper introduce Standard 3D object(Bounding-Volume). In 3D game very efficient control algorithm Using collision detection which controls the convenient of a game based on Standard 3D object specially collision-detection. This algorithm is designed LOD(Level-of-Detail) using of collision detection efficiency improvement in 3D Object Furthermore, we also show how our method lends itself naturally for an adaptive 3D graphics and a realistic expression method, which may be of critical importance in some applications. Consequently, we propose two different techniques for collision detection in 3D, one is using ‘LOD’ our hierarchical data structure, and the other is an improved efficient realism ‘realism’ based method.

This paper presents a parallel kd-tree traversal algorithm based on the parallel binary radix tree construction scheme proposed by Tero Karras in 2012. In his paper, Tero Karras proposed parallel tree construction algorithm which can maximize the utilization of GPU threads, but implementation and analysis of kd-tree are not fully discussed. This paper aims to fill the gap for the specific kd-tree cases. As an application for the kd-tree traversal method proposed in this paper, the proposed method has been implemented with NVIDIA’s CUDA framework and tested on NVIDIA’s realtime raytracing library, OptiX. As a result, the proposed method can construct tree structures within the requirement for realtime process, but still needs specialized spatial caching data structure like the “primitive tree” for highly detailed meshes to handle spatial queries as fast as to visualize implicit surfaces under OptiX framework. However, the proposed method can be applied to dynamic collision detection and manage scene and object information for game applications thanks to its fast tree construction process. Also, realtime raytracing can be applied to game applications based on this study.

With the development of the Internet technology, people’s life and work are more and more dependent on the Internet. Remote rendering techniques permit streaming of high-quality 3D graphics to apply onto a wide range of smart devices, and recent years have also seen many solutions of Web3D applications. Web3D is a kind of technology, which can perform the 3D visualization on the Internet, and virtual roam can present information in a far more intuitive and interactive way to people. So, Web3D technology is the foundation of virtual roam technology. Today, Web3D technology and virtual roam technology has become the research focus of computer technology. This paper presents the basic concept and characteristics of Web3D, and designs a solution concerning the virtual roam by combining WebGL and HTML5. This solution has characteristics of good compatibility, GPU acceleration and free plug-ins. Meanwhile, this paper aims at researching collision detection in the virtual roam and presenting a fast collision detection and correction algorithm.

The hand interaction between a virtual object and human on Augmented Reality is a major issue to manipulate a rendered object in a convenient way. Conventional 2D image-based recognition and interaction technique in AR has a limitation to perform a natural interaction between the user and the virtual object. In this paper, we present a natural 3D hand interaction with the augmented object by the use of stereo-vision technique. The proposed 3D hand interaction technique can obtain 3D hand position and finger direction while hand is approaching to the virtual object. 3D hand interaction can manipulate the virtual object in a feasible fashion. In addition, it can provide various interactions by detecting the collision between the 3D hand and the virtual 3D object on AR.

The objective of this work is to propose two new algorithms for collision detection for real-time application. They are applicable to rigid objects enclosed in boxes in order to improve the time of collision detection. The proposed algorithms, called Neuro-SAT and perceptron learning with displacement of the base frame, will be compared with the algorithm Separating Axis Test (SAT) based on the hierarchy of the OBB tree. A grasping operation of an object, with a robotic hand, was executed to test the developed algorithms. The results, of simulation experiments, reveal a great improvement in the time of collision detection.

This paper presents a novel deep learning-based radio frequency identification (RFID) tag collision detection method for ultra-high frequency (UHF) RFID. UHF RFID technology provides longer communication range compared to NFC, barcode, and QR code technology. However, due to the longer range, multiple tags in wide range may reply simultaneously such that a reader receives superposed signal of multiple tags. Multiple tag signals interfere with each other such that reader’s tag reading speed is decreased. In order to detect these tag collisions, previous studies utilized analytical methods rather than theoretical ones. Hence, a deep learning-based solution can improve the detection performance. For deep learning, training datasets are generated from mathematical equations, which are specified by the standard, with various delay times, amplitude differences, phase differences and noise level among tag signals. Arbitrary delay time, phase difference, and amplitude difference are used in every run of simulation. Simulation results show that the detection performance using the proposed method is about 5 dB better than that of existing method.

War State Battle Field is the basement of the military action. Battle Field is very significant for commanders to considerate and make full use of it in decision-making. This is basically depend on the vehicular traffic control in war state battle field. In this the tanks represents a vechicle. A tanks traffic on war state battle field is a vital problem and is seemly a major pretend to conclusion makers. In this research paper we will try to bring the scenario of battle field arena where the traffic consists of the vehicles called tanks. So the main focus in this research is put on the collision detection of the tanks among each other using cloud sensor by controlling the congestion in the battlefield.

Advances in MEMS (Micro Electro Mechanical Systems), microprocessor, and wireless technologies, have accelerated deployment of network-capable computing devices attached to various objects over the world. These devices should be small in size and have a characteristic of extremely low power consumption. Therefore, in 2004, IEEE 802.15.4 has been established as a global standard for LR-WPANs. After that, the phenomenal popularity of this standard has also led to indiscreet deployment of other WPANs. Therefore, LR-WPANs can suffer from collision by other WPAN devices. In order to solve collision problem in IEEE 802.15.4, in this paper we introduce a novel channel scheduling scheme which is an adaptive method with fast recovery. The proposed scheme can minimize the possibility of beacon collisions by efficiently managing the multiple available channels in a hybrid manner combining proactive and reactive methods.

This paper proposes an efficient collision detection method in the dynamic spatial subdivisions for the MMORPG engine which requires realtime interactions. An octree is a suitable structure for static scenes or terrain processing. An octree spatial subdivision enhances rendering speed of scenes. Current spatial subdivisions tend to be highly optimized for efficient traversal, but are difficult to update quickly for a changing geometry. When an object moves to the outside extent for the spatial subdivisions, the acceleration structure would normally have to be rebuilt. The OSP based on a tree is used to divide dynamically wide outside which is the subject of 3D MMORPG. TBV does not reconstruct all tree nodes of OSP and has reduced rebuilding times by TBV information of a target node. A collision detection is restricted to those objects contained in the visibility range of sight by using the information established in TBV. We applied the HBV and ray tracing for an efficient collision detection.

This paper presents a real-time algorithm for collision detection, collision avoidance and guidance. Three-dimensional point-mass aircraft models are used. For collision detection, conflict probability is calculated by using the Monte-Carlo Simulation. Time at the closest point of approach(CPA) and distance at CPA are needed to determine the collision probability, being compared to certain threshold values. For collision avoidance, one of possible maneuver options is chosen to minimize the collision probability. For guidance to a designated way-point, proportional navigation guidance law is used. Two scenarios on encounter situation are studied to demonstrate the performance of proposed algorithm.