摘要
运用CNN设计了一套生物芯片样点识别算法。算法实现的目标:改善已有方法的缺陷,达到良好的图像质量增强效果;将CNN输出的模拟信号图像转化为样点数据信息,使得后续的信息分析成为可能。最后利用实际CNN芯片参数估算了整套算法的运算时间,结果显示其速度达到实时处理的标准。
Based on Cellular Neural Networks (CNN), an algorithm for spots identification of microarray image was designed in this paper. The proposed algorithm achieves two goals:modified the algorithm used recently, results show that enhancement of image quality can be done by our algorithm; transformed the analog output image of CNN to digital data, which made the consequent information analysis possible. Finally, we estimate the running time of the algorithm by taking into account the parameters of a real CNN chip, and the results show that the speed of our algorithm achieves the standard of real-time processing.
出处
《电子技术应用》
北大核心
2008年第5期127-129,133,共4页
Application of Electronic Technique
基金
国家自然科学基金资助项目(基于形态小波和细胞神经网络算法的生物芯片样点识别新方法)(60671046)