图像目标检测算法研究

论述了一种图像目标检测算法。利用区域PCA对图像进行变化,用低维子空间描述高维空间中的图像。将低维子空间中的向量加载到BP网络的输入端进行训练,调整神经网络权值。然后利用建立起来的BP网络对图像中的目标进行检测;并且在此过程中,建立多种训练的目标库和背景库。最后编写了一个实现整个系统的软件,将图像输入软件,根据神经网络检测,输出端即可得到识别结果。

Target Detection Arithmetic

The detection of functional images has been discussed.PCA transform has been used to extract featrues and describe the hight-dimensional space were loaded onto the input layer of BP n et work and started training.Weights were adjusted until a stable status was reached.BP n et work was used to detect target from the image.And in this course,a target database and a background database had been constructed.And when preprocessing intensity images to be recognized were loaded onto the input layer of BP network,recognition results were obtained at output layer. [