solution

用灰度共生矩阵提取五个纹理特征 包括：熵、均匀度、相关性、能量、对比度(Extraction of five GLCM texture characteristics，Include: entropy, uniformity, correlation, energy, contrast)

gmsk调制程序，matlab仿真程序，可以叠加程序(gmsk transform program, matlab simulation program, the program can be superimposed)

结构可靠度分析模型拟合样本点数目影响下的功能函数画图，分别采用ANN和SVM模型(Structural reliability analysis model fitting function drawing under the influence of the number of sample points, respectively, ANN and SVM models)

一个bayes 算法的face recognisation的matlab code，是我的一个作业(A Bayes algorithm face recognisation the matlab code, is one of my operation)

emd能量熵的matlab代码................(The emd energy entropy of matlab code................)

1 录入一段语音，对其进行时域频域处理，通过变频、变速达到变声目的，且使处理后的语音信号不影响正常收听与理解。 2 用图形用户界面（GUI）对上述变声前变声后进行时域频域图形输出，不同的频率、速度有不同的输出，且对前后波形进行比较 (1 entry a voice, the frequency domain and time domain by frequency conversion variable speed, processing to make the treatment and voice purpose, after the speech signal shall not affect the normal to listen to and understand. 2 with the graphical user interface (GUI) voice of the above voice before after the frequency domain in the time domain graphics output, different frequency, speed have different output, and compared before and after waveform )

simulink通信系统仿真，频分多址，时分多址，码分多址(simulink simulation of communication systems, frequency division multiple access, time division multiple access, code division multiple access)

遗传算法和神经网络联合用于数据融合，附说明文档和参考文献，matlab应用(Genetic algorithms and neural networks for data fusion joint, with documentation and references, matlab application)

模糊聚类的改进--先用减法聚类得到初始点，而后用模糊聚类得到聚类中心，能够提高聚类的速度和精度(Improved Fuzzy Clustering- first get the initial point subtraction clustering, and then get the cluster center with fuzzy clustering, clustering can improve the speed and accuracy)

本文主要做了1.对三维重建系统中的预处理算法和分割算法进行细致研究，并通过大量实验进行对比分析，指出其优缺点及适用围。 2.传统的分割算法处理每一幅图片都需要用户选定对象和背景，耗时较长。针 对医学CT连续断层图像间相关性强特点，提出一种把Graph Cut方法和区域生长方法相结合的图像分割算法。3.最后将分割目标与后续重建算法结合起来，建立一个完整的三维重建系统。 (This paper has 1 of 3D reconstruction system of preprocessing and segmentation algorithm is carefully studied, and through a lot of experiments are compared and analyzed, and points out their advantages and disadvantages and application range. 2 traditional segmentation algorithm to deal with each picture requires the user to select the object and background, time-consuming. For medical CT images correlation strong characteristic, put forward a segmentation algorithm combining Graph Cut method and region growing method. 3 finally the target segmentation and subsequent reconstruction algorithms are combined, the establishment of a complete 3D )