论文部分内容阅读
在对智能整骨机器人中的骨折X线图像的特点和噪声来源进行分析的基础上,提出了利用小波变换的多分辨分析特性,采用时间窗和频率窗都可以改变的时频局部化方法,使图像内的有用信号集中在小波变换的少数系数上,而噪声信号集中到其他小波系数上。在提高图像信噪比的同时,保留了图像的位置信息,确保了临床中利用该智能整骨机器人进行基于模糊控制和神经网络的骨折自动诊断和自动整复的准确度。
Based on the analysis of the characteristics of the fractured X-ray images and the sources of noise in the intelligent osteo-robot, a multi-resolution analysis using wavelet transform is proposed. The time-frequency localization method, which can change both time window and frequency window, The useful signals in the image are concentrated on a few coefficients of the wavelet transform, while the noise signals are concentrated on other wavelet coefficients. While improving the signal to noise ratio of the image, the position information of the image is preserved, which ensures the accuracy of automatically diagnosing and automatically reconstructing fractures based on the fuzzy control and the neural network in the clinic.