| 超薄精密件对接接头图像采集和信息识别 |
| |
| 引用本文： | 聂大涛,何建萍.超薄精密件对接接头图像采集和信息识别[J].应用声学,2015,23(6). |
| |
| 作者姓名： | 聂大涛 何建萍 |
| |
| 作者单位： | 上海工程技术大学材料工程学院 |
| |
| 基金项目： | 国家自然科学基金项目（面上项目，重点项目，重大项目） |
| |
| 摘 要： | 针对0.1mm不锈钢超薄板的超薄和微小尺寸等特点，研究其对接接头图像采集方案和信息识别方法。对比分析了曝光时间、亮度对图像采集的影响，采用Levenberg-Marquardt算法和全局优化法得出曝光时间、亮度和图像平均灰度值三者之间的匹配关系式。为了消除对接接头边缘毛刺、划痕和颗粒噪声对接头图像信息识别的影响，运用阈值分割、腐蚀和膨胀等形态学处理方法进行“去噪”处理，精确的得出了超薄板对接接头的信息
|
| 关 键 词： | 超薄板 对接接头 图像采集 信息识别 |
Image acquisition and information identification of ultra-thin precision parts butt joint |
| |
| Abstract: | Aiming at micro-dimension characters of ultra-thin stainless steel sheet with 0.1 mm thickness, image acquisition and information recognition of the butt joints were studied. The effects of exposure time and brightness on image acquisition were analyzed. Matching criterion, among the exposure time, brightness and average gray value, was obtained using Levenberg-Marquardt algorithm and global optimization method. In order to eliminate the influence of butt joint edge burr, scratches and granular noise on the image information recognition of joints, image processing methods, such as threshold segmentation, erosion, and dilation to de-noising, were inducted to obtain exact butt-joint information of ultra-thin sheets. |
| |
| Keywords: | ultra-thin sheet butt joint image acquisition information recognition |
|
| 点击此处可从《应用声学》浏览原始摘要信息 |
| 点击此处可从《应用声学》下载免费的PDF全文 |
|
