The Extracting System Design of Small Dim Targets in Complex Background

Dong Sheng Liang; Zhao Hui Liu; Wen Liu

doi:10.4028/www.scientific.net/AMR.328-330.2324

Paper Titles

Buffer and Vibration Optimization of Missile Data Recorder Structure
p.2304

Design of Synthetic Diagnostic and Health Management of Launch Vehicle
p.2308

The Method and Application of Time Series Prediction Based Wavelet Neural Network
p.2312

Design Procedure of Robust QFT-Based Controller for Continuous-Flow Grain Dryer Plant
p.2318

The Extracting System Design of Small Dim Targets in Complex Background
p.2324

Based on Rotating Machinery Data Acquisition System Anti-Aliasing Filter Design and Application
p.2328

Study on Integrated Setup System with Computer Vision for Breast Cancer Radiotherapy
p.2333

ONAMA: A Quantum Leap in High Performance Computing
p.2337

Optimization of Comprehensive Assessment Parameters System for Equipment Maintenance Quality Based on Principal Component Analysis
p.2343

HomeAdvanced Materials ResearchAdvanced Materials Research Vols. 328-330The Extracting System Design of Small Dim Targets...

The Extracting System Design of Small Dim Targets in Complex Background

Abstract:

For the moving small dim targets in visible image sequences with low SNR and complex background, whose contained characters are simple and poor, they are extracted difficultly. This paper proposes a new method to detect and extract positions of small dim moving targets. According to the features of moving targets, in a very short time interval, the target trajectory is considered as a straight line approximately. Firstly, it makes use of threshold segmentation methods to extract the positions of targets in each frame, then building the motion line equations after the joint of multi-frame processing results. Finally, the position of small dim targets are detected out and extracted, and false targets are eliminated accurately. Hardware system was designed and the algorithm is implemented on hardware systems successfully. The experiment results show related functions of the system and extracting algorithm is feasible, the system is stable and has a strong process ability, which can effectively detect and extract small and dim target in complex background correctly.