成像空间分辨率均衡化与目标快速检测技术

FAST TARGET DETECTION TECHNOLOGY OF EQUALIZING RESOLUTION IMAGE SPACE

分析了红外玫瑰线扫描亚成像制导的成像空间分布函数 ,提出一种基于神经网络模型的成像空间分辨率均衡化目标快速检测技术 .该技术针对亚图像成像空间分辨率的非线性特性 ,有效利用每个像素所携带的信息 ,实现亚图像目标的均衡识别 ,从而提高系统的全域检测概率 .实验结果表明该技术方案是可行的 。

Through the analysis of the distribution function of the rosette scan infrared subpixel seeker, a fast target detection technology of equalization of resolution of image space was proposed. In consideration of the nonlinear characteristics of the image space, by making full use of the information of each pixel, this technology realizes the subpixel target recognition, and improves the target detection probability. The experimental results show that this method is feasible and can be realized in real time in DSP system.