원문정보
초록
영어
To solve the problem of moving object detecting in the wide field, this paper establishes a new artificial compound imaging system and proposes moving object detection algorithm. According to the unique principle of receiving and processing of visual information from animal compound-eye, the model of artificial imaging system consists of a micro-lens array and CCD device with planar array, captures dynamic image sequences in the different azimuth, and assures its advantages of long-distance and wide-range detection and good image effect. In order to get the motion vector of moving object, researches the characteristics of background in the relevant imaging sequence, discusses the mechanism of processing visual information, the estimation algorithm adopt the high-pass filter processing after making the three-dimensional space transformation to reduce the disturbance of noise and enhance the effect of background suppression firstly. The motion vectors of dynamic object is obtained by processing the relevant imaging sequence with the maximum correlation coefficient, refers to the EMD array. Preliminary experimental results for target tracking shown that the artificial compound imaging system and moving object detection algorithm is practical and effective, and which has a highly practical application value.
목차
1. Introduction
2. Insect Visual Target Detection Mechanism
2.1. The Physical Structure of Fly Visual System
2.2 Target Detection Model of the Fly’s Nervous System
3. Imaging System of Artificial Compound-Eye
4. Moving Target Detection Simulation
4.1. Image Acquisition and Image Segmentation
4.2. Background Suppression with High-Pass Filter and Morphological Filter
4.3. Motion Vector Information Processed by EMD
5. Conclusions
References
