木材缺陷智能检测技术可以有效降低人工误检带来的经济损失,对提高木材加工智能化水平具有重要意义。提出了一种木材缺陷智能检测算法,通过双循环生成对抗网络(double least generative adversarial networks,DLGAN)及密集卷积网络(Dens...木材缺陷智能检测技术可以有效降低人工误检带来的经济损失,对提高木材加工智能化水平具有重要意义。提出了一种木材缺陷智能检测算法,通过双循环生成对抗网络(double least generative adversarial networks,DLGAN)及密集卷积网络(Dense-Net)来检测色差、虫眼、裂纹、节子和伤疤等5种木材常见缺陷。首先,使用DLGAN技术扩充数据集,提高数据集的多样性和数量,缓解了因训练数据不足而导致的过拟合问题;其次,基于Dense-Net的特点,采用密集的卷积块序列提高对微弱特征的提取和学习能力,以便更好地检测木材缺陷。试验结果表明,相比VGG16、Inception-v2、ResNet 3种经典卷积神经网络,基于DLGAN增广数据集训练的Dense-Net模型有效提高了木材缺陷检测模型的性能,平均准确率达到92.7%,在只使用少量训练数据的情况下模型依然具有良好的图像生成能力和训练鲁棒性。展开更多
The paper first describes the watershed algorithm and solves the problem of the over-segmentation from the watershed algorithm by using the mark watershed transform; then wood defect image is treated with algorithm; f...The paper first describes the watershed algorithm and solves the problem of the over-segmentation from the watershed algorithm by using the mark watershed transform; then wood defect image is treated with algorithm; finally the comparison is made between the original image and the edge image detected. The result showed that the wood defect image could be segmented with the mark-controlled watershed this algorithm and the defect edge image be exactly detected. Moreover, the treatment also provided the convenience for the following treatment such as pattern recognition.展开更多
文摘The paper first describes the watershed algorithm and solves the problem of the over-segmentation from the watershed algorithm by using the mark watershed transform; then wood defect image is treated with algorithm; finally the comparison is made between the original image and the edge image detected. The result showed that the wood defect image could be segmented with the mark-controlled watershed this algorithm and the defect edge image be exactly detected. Moreover, the treatment also provided the convenience for the following treatment such as pattern recognition.
