引入置信区间图谱探析系统优化策划.docVIP

引入置信区间的图谱分析系统优化设计 　　摘 要： 当前的图谱图像分析系统通过图像特征高速峰值进行特征检测与分割，检测到的图谱图像脉冲峰值具有不可预测性，系统受到的干扰较大，容易产生图谱图像特征泳道粘连干扰，检测效果不佳。提出一种小型化、低功耗和全波采样的新型的图谱图像分析系统优化设计方案，引入一种置信区间的理论，根据图谱特征背景灰度值的大小计算谱带，运用高斯分布概率设定一个置信区间，运用置信区间对粘连泳道进行最优分割处理，完成分析。仿真结果表明，以DNA指纹图谱图像为例，采用该系统能有效实现对图谱进行分析，测谱准确，有效抑制了干扰失真，且功耗较低，展示了其优越性能 关键词： 图谱分析； 灰度特征； 系统设计； 置信区间 中图分类号： TN911?34 文献标识码： A 文章编号： 1004?373X（2015）24?0023?04 Optimization design of image spectrum analysis system introducing confidence interval SUI Dan1， 2， HE Fang2 （1. School of Information Engineering， Wuhan University of Technology， Wuhan 430070， China； 2. School of Software， Anyang Normal University， Anyang 455000， China） Abstract： The high?speed peak value of image feature is adopted in the current spectrum image analysis system to conduct feature detection and segmentation， in which the detected pulse peak value of the spectrum image has unpredictability， and the system is influenced by the interference greatly and prone to cause lane adhesion interference of the spectrum image feature， so it results in poor detection effect. An optimization design scheme applied to the new spectrum image analysis system of miniaturization， low power consumption and full?wave sampling is proposed. The theory of confidence interval is introduced to calculate the band according to the grey value of the spectrum feature background， and the confidence interval is set with Gaussian distribution probability， which can be used to conduct the optimal segmentation of the adhesion lane， so the analysis is completed. The simulation results show that， taking DNA fingerprint image as an instance， the system can analyze the image spectrum effectively， which is accurate for spectral measurement， can restrain the disturbance distortion effectively， and has low power consumption. The superior performance of the system is showed. Keywords： image spectrum analysis； gray characteristic； system design； confidence interval 0 引 言 对各种特征进行图谱分析，在很多领域有着广泛的应用，其中以图像处理为