一种基于大尺寸的不平衡样本的焊缝图像识别分类算法

• Format: Patent
• Language: Chinese
• Published: 07.04.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

本发明公开了一种基于大尺寸的不平衡样本的焊缝图像识别分类算法，本发明计算好焊缝与缺陷焊缝的相似度，找出与缺陷焊缝图像相似度最高的好焊缝图像，用好焊缝与该缺陷焊缝图像进行图像融合生成新的焊缝缺陷图像，再输入生成对抗网络，进行充分的对抗训练之后，可生成指定缺陷的缺陷样本，本发明通过先融合再生成的算法可以避免生成自由度过大的样本，通过该算法扩充数据集后，极大的减小因为类不平衡导致对缺陷分类问题的影响，提高识别算法的精度；本发明算法不仅可用于焊缝的缺陷图像数据集，同时可以用于其他正品样本与缺陷样本之间差距不大的情况下生成新的缺陷样本。 The invention discloses a welding seam image recognition and classification algorithm based on a large-size unbalanced sample, which comprises the following steps: calculating the similarity between a good welding seam and a defective welding seam, finding out a good welding seam image with the highest similarity with a defective welding seam image, and carrying out image fusion on the good welding seam and the defective welding seam image to generate a new welding seam defect image. A sample with an overlarge degree of freedom can be prevented from being generated through an algorithm of first fusion and then generation, after a data set is expanded through the algorithm, the influence on the defect classification problem caused by class imbalanc