仪表表盘数字识别方法的研究

Research on Number Recognition Method of Instrument Dial

针对无指针式仪表表盘的数字识别问题,提出一种基于特征提取和粗糙集特征约简的神经网络数字识别方法。该方法首先利用数字图像预处理技术处理图像并利用特征提取方法提取数字图像特征,然后利用粗糙集理论进行特征约简,最后将约简后的信息输入到训练好的神经网络进行识别。

The paper put forward a neural network number recognition method based on feature extraction and rough set feature reduction aiming at number recognition problem of no-pointer instrument dial.Firstly the method processed images using pre-process technology of number image and extracted number image features using feature extraction method,then carried through feature reduction using rough set theory,finally carried through recognition with reduced information into trained neural network.