猛犸象牙与现生象牙的计算机视觉鉴定方法

Computer vision-based identification method for mammoth and live ivory

文章提出一种基于计算机视觉的象牙鉴定方法,通过专用图像采集设备实现高清象牙纹理图像拍摄,采用图像增强方法对原始图像进行预处理并建立象牙纹理图像数据集,基于深度神经网络训练象牙识别模型对象牙纹理图像进行特征提取和分类,从而实现猛犸象牙与现生象牙鉴定。结果表明该方法能够快速、有效地对象牙制品进行鉴定与识别,解决目前的执法难题。

In this study,a computer vision-based ivory identification method is proposed.A special image acquisition device is used to capture high-resolution ivory texture images,a dataset of ivory texture images is established,and deep neural networks are used to train ivory recognition models to extract and classify ivory texture images,so as to realize the identification of mammoth ivory and live ivory.The experimental results show that this method can quickly and effectively identify ivory products and solve the current law enforcement problems.