基于双目视觉与Transformer的连铸坯模型定位与测量OA北大核心CSTPCD
Continuous casting slab model positioning and measurement based on binocular vision and Transformer
针对双目视觉传统检测算法效率低、匹配复杂等问题,本文提出一种基于双目视觉与Transformer的连铸坯模型定位与测量方法.首先,使用标定后的平行双目相机采集连铸坯模型左右图像,经校正、标注后将其作为数据集.然后以改进的Transunet*为骨干,利用神经网络对数据集输出关键点坐标,网络模型采用多尺度U型结构来抵消因下采样量化而产生的高斯热图理论误差下界.为改善卷积神经网络只关注局部特征的缺陷,加入Transformer结构来强化每个通道内的信息…查看全部>>
In order to address the problems of low efficiency and complex matching of traditional binocular vision detection algorithms,a continuous casting slab model positioning and measurement based on binocular vision and Transformer method was proposed in this paper.Firstly,a calibrated parallel binocular camera was used to collect images of the continuous casting slab model,which were used as datasets after correction and labeling.Then,with the proposed Transunet…查看全部>>
李同谱;许四祥;施宇翔;杨利法
安徽工业大学机械工程学院,安徽马鞍山,243032安徽工业大学机械工程学院,安徽马鞍山,243032安徽工业大学机械工程学院,安徽马鞍山,243032安徽工业大学机械工程学院,安徽马鞍山,243032
计算机与自动化
双目视觉Transformer关键点检测注意力机制
binocular visionTransformerlandmark detectionattention mechanism
《中南大学学报（自然科学版）》 2024 (4)
1312-1322,11
国家自然科学基金资助项目(51374007)安徽高校自然科学研究重点项目(KJ2020A0259)Project(51374007)supported by the National Natural Science Foundation of ChinaProject(KJ2020A0259)supported by the Key Project of Natural Science Research of Anhui Educational Committee)
评论