人工智能深度学习的激光图像分割研究

Research on laser image segmentation based on artificial intelligence deep learning

激光成像受到环境、设备自身等干扰,使得激光图像含有噪声,当前图像分割方法对噪声干扰鲁棒性差,误分割现象出现概率高,重要信息丢失严重,为了克服当前激光图像分割的弊端,提出了基于人工智能深度学习的激光图像分割方法。首先采用小波变换对激光图像进行特征提取,并对噪声干扰进行抑制处理,然后引入人工智能学习算法对激光图像特征向量进行训练,并根据训练结果对激光图像像素点进行分类,从而实现激光图像分割,最后采用含噪和不含噪的激光图像进行仿真测试。结果表明,对含噪和不含噪的激光图像,人工智能深度学习的分割精度分别达到了91%和95%以上,精度明显高于经典激光图像分割方法,分割效率可以满足激光图像向大规模方向发展的要求。

The laser image is interfered by the environment and equipment itself,making the laser image contain noise.The current image segmentation method has low robustness to noise interference,high probability of false segmentation and severe loss of important information.To overcome the disadvantages of current laser image segmentation,an artificial intelligence deep learning laser-image segmentation method is proposed.Firstly,the wavelet transform is used to extract the laser image features,and noise interference is suppressed.Then an artificial intelligence learning algorithm is introduced to train the feature vector of laser image,and the pixels of laser image is classified according to the training results to realise laser image segmentation.Finally,the laser image with noise and without noise is used for the simulation test.The results show that for the laser image with and without noise,the segmentation accuracy of artificial intelligence depth learning reaches 91% and 95% respectively,which is significantly higher than that of the classical laser image segmentation methods,and the segmentation efficiency can meet the requirements of large-scale development of laser image.