Earticle

광학시스템에서 가장 중요한 기술 중의 하나는 고속으로 이동하는 표적을 지속적으로 추적할 수 있는 고속 자동 표적 추적 시스템이다. 본 논문은 원거리 소형 표적에 대해서 고속으로 이동하는 표적에 대해서 급격한 형태 변화에도 강인한 상관 트렉커 기반에 자동 표적 추적 시스템을 설계한다. 본 논문은 IR 영상에서 3x3 이상의 표적에 대해서 4ms 내에 고속으로 표적을 추적하기 위한 커널 함수와 correlation filter 설계 최적화 방법을 제시하고, 이를 VxWorks와 같은 실시간 O/S 하에서 짐벌과 함께 연동하여 시험을 수행한다. 제안된 알고리즘의 성능 검증을 위해서 실제로 복잡환 경에서 기동하는 드론, 비행체를 대상으로 IR 카메라로 영상을 획득한 후 이를 실시간 보드상에서 시험을 진행한 결과 응답시간 4ms이하에서 약 98% 추적 성공률를 보였다.

On today, optical system are the next generation weapon systems being studied in many countries, starting from USA. One of the most important technologies in optical system is a high-speed automatic target tracking system that can continuously track high-speed moving small targets. This paper designs an automatic target tracking system based on a correlated trekker that is robust against rapid shape changes for fast moving targets and small targets at a distance. The proposed system showed about 98% success rate in response to the targets that are under a complex background such as drone, ranger, etc.

An active camera tracking system (ACTS) is comprised of a camera that can rotate automatically toward an object of interest so the object remains within the camera’s field of view. To rotate the camera toward an object, the camera needs to detect the object’s location. Detecting an object’s location can be achieved using image processing techniques. However, conventional image processing techniques require expensive calculations in addition to expensive hardware, which in turn limits the video frame processing speed. In this paper, we propose an effective ACTS to monitor and track objects automatically with high accuracy. The proposed approach automatically initializes the object to be tracked using an absolute difference motion detection technique. To track the object correctly, the minimum output sum of squared error (MOSSE) adaptive correlation tracker is applied to provide an accurate and fast visual tracking. We also implemented an accurate scale estimation for the visual tracking technique that detects the scale changes of the object. Additionally, the proposed ACTS uses a closed loop controller to control the pan-tilt speed of the system. By combining these techniques, the proposed ACTS can provide accurate tracking with smooth rotation.

This paper propose a target tracker using phase correlation. The tracker consist of a pre-processing module, a translation estimation module based on phase correlation, a fine motion estimation module applied when confidence rate could not fulfill a threshold value and a reference image update module. The fine motion estimation module measure the shift, rotation and scale of input image compared to reference using Fourier-Mellin transform. Proposed tracker was tested its accuracy and robustness using some real indoor and outdoor image sequences.