Enhancement of Physical Modeling System for Underwater Moving Object Detection

이동하는 수중 물체 탐지를 위한 축소모형실험 시스템 개선

Abstract

Underwater object detection method adopting electrical resistivity technique was proposed recently, and the need of advanced data processing algorithm development counteracting various marine environmental conditions was required. In this paper, we present an improved water tank experiment system and its operation results, which can provide efficient test and verification. The main features of the system are as follows: 1) All the processes enabling real time process for not only simultaneous gathering of object images but also the electrical field measurement and visualization are carried out at 5 Hz refresh rates. 2) Data acquisition and processing for two detection lines are performed in real time to distinguish the moving direction of a target object. 3) Playback and retest functions for the saved data are equipped. 4) Through the monitoring screen, the movement of the target object and the measurement status of two detection lines can be intuitively identified. We confirmed that the enhanced physical modeling system works properly and facilitates efficient experiments.

최근 전기비저항 탐사의 정밀 계측기술을 활용한 수중 물체 탐지방법이 제시되었고, 변화하는 해양환경에 대응할 수 있는 자료처리 기술 고도화 연구의 필요성이 제기되었다. 이 연구에서는 효율적인 실험과 검증을 위한 개선된 축소모형실험 시스템과 그 운용 결과를 제시한다. 이 시스템은 다음과 같은 특징을 가진다. 1) 실시간 실험영상과 계측자료의 동시 수집 및 분석과 같은 모든 프로세스가 5 Hz의 속도로 이루어진다. 2) 두 개 탐지선 자료의 실시간 계측 및 처리로 수중물체의 이동방향 파악이 가능하다. 3) 저장된 자료를 이용한 반복실험이 가능하여 획득된 자료의 다각도 반복분석이 가능하다. 4) 모니터링 화면을 통해 수중물체가 이동하는 모습과 두 탐지선 자료를 동시에 직관적으로 파악 가능하다. 개선된 시스템을 이용한 실험 결과, 모든 시스템이 정상 작동하고 효율적 실험이 가능함을 확인하였다.

Fig. 2. Comparison of block diagrams between the previous by Cho et al. (2016) (a) and the enhanced version presented in this study (b) of physical modeling systems. The main features of new system are as follows: All processes are performed at 5 Hz refresh rates rather than 3 Hz. A camera is equipped to record the situation of experimental conditions. Two detection lines are used to determine the direction of the moving target object. The number of input channel of data acquisition unit increases. In the data processing and control unit, two background mode and test mode are available to retest the stored data.

Fig. 1. Conceptual installation of the underwater object detection method (Cho et al., 2016). Two detection lines are located on the seabed and each detection line consists of two current electrodes and a number of potential electrodes. C1-C4 represent current electrodes.

Fig. 3. Photograph of the enhanced version of physical modeling system. The system consists of data processing and control unit, camera, power supply, target object, two detection lines, and data acquisition unit.

Fig. 4. The screen configuration of the enhanced physical modeling system. Panel A: video image is displayed to understand the experimental conditions. Panel B: control and monitoring panel is placed. Panel C and D: 2D and 3D graphs of detection line 2 and 1, respectively. In 2D graph, real-time data are displayed and in 3D graph, stacked real-time data are displayed.

Fig. 5. Flowchart of the video monitoring process.

Fig. 6. Flowchart of the data processing routine.

Fig. 7. Flowchart of the mode 3 (background data generation algorithm).

Fig. 8. Plan and sectional views of the water tank experiment. Eachdetection line consists of two current electrodes at both ends and24 potential electrodes with 1-cm spacing. The stimulation currentlevel is set to 6mA and electric field data for 23 consecutive pairsof adjacent potential electrodes are acquired for each detection line.The target object is moving along the pathway between potentialelectrodes 6 and 7.

Fig. 9. Results of the water tank experiments using mode 1 (without background data) at the detection line 1 (a) and the detection line 2(b). In the 3D graphs, the x-axis represents the channel number, the y-axis represents the time in seconds, and the z-axis represents thedistortion level of electric field.

Fig. 10. Results of the water tank experiments using mode 2 (background data update algorithm) at the detection line 1 (a) and the detection line 2 (b).

Fig. 11. Results of the water tank experiments using mode 3 (background data generation algorithm applied) at the detection line 1 (a) and the detection line 2 (b).

Fig. 12. Console displays at the moment when the target object is passing the right side of water tank (a), the right above the detection line 1 (b), between detection line 1 and 2 (c), the right above the detection line 2 (d), respectively. Note that drastic changes of electric field distortion are observed according to the position of the target object.

Table 1. Summary of the previous system by Cho et al. (2016) and the enhanced system.