光伏阵列故障状态红外图像的分割研究与实现

Research and Implementation of Segmention for Fault States of Photovoltaic Array Infrared Image

依据光伏阵列故障状态红外图像的特点,提出了一种高效基于迭代的两级阈值分割算法。实验表明:该算法能较好地分割出光伏阵列中存在的正常、故障、老化三种区域,同时算法以乘法和加法运算为主,便于DSP的高速运算。最后,采用TMS320 DSP算法标准(XDAIS)对分割算法进行创建与封装,极大地方便了系统集成。

According to the characteristics of photovoltaic array infrared images,the paper proposed an effective two-level threshold segment method based on iteration.The experimental results indicate that the normal,fault and aged three regions of photovoltaic array could be segmented more correctly,at the same time,the algorithm using multiplication and addition operations is convenient for high-speed operation on DSP.Finally,the segmentation algorithm based on the TMS320 DSP Algorithm Standard（XDAIS） is created and packaged,which greatly facilitates system integration.