基于CNN的海空光电目标检测技术研究

Research of sea-aero target detection from photoelectricity image based on cellular neural networks

根据元胞神经网络图像处理模型,提出了一种海空光电目标的检测方法。与传统的方法不同,元胞神经网络是一种新的图像处理模式。它把图像看成一个动力系统,如果设置合适的模板参数与初始条件,就能达到特定的图像处理功能。首先对图像进行预处理和直方图修正,然后利用CNN进行分割,最后检测出目标。通过仿真实例,结果表明,该方法能较好地检测出海空目标,准确率较高,并且满足实时处理的要求,具有一定的军事应用价值。

According to Cellular Neural Networks model for image processing, a method is proposed appropriate for sea-aero target detection from photoelectricity image. The difference from traditional method, CNN is a new image-processing pattern. It regards the image as a dynamic system. If appropriate templates and initial condition are set, it can realize specific function. After pre-process and histogram altering, it utilizes CNN to realize segmentation, and finally the targets are detected. The simulation experim...