针对传统仓库管理系统中螺丝钉统计计数难题,将数字图像处理引入到螺丝钉检测计数中.提出了一种基于图像处理的螺丝钉自动计数方法.首先对未放置螺丝钉的货架进行处理,检测并记录放置螺丝钉的位置等信息,建立模板.在建立模板的过程中,采用人机交互的方式进行校正,能够有效地解决光照、噪声等带来的影响,从而建立准确的模板.然后将未放置螺丝钉的图像(模板图)与待检测图像相减,突出放置螺丝钉后的位置的特征.最后采用模板匹配法,检测对应位置是否存在螺丝钉.实验表明：该方法能够有效地解决货架受损、螺丝钉氧化和噪声干扰等情况带来的影响,并且实现简单.计数准确率高达98％以上,平均检测时间为3．67 ms,很大程度地提升了仓库管理性能.
In view of screws counting problem in traditional warehouse management system,digital image processing is introduced to the detection and counting of screws.A method for screws automatic counting based on image processing ispresented.Processing the shelves without placing screws is to detect and record placed screw position information,establishing the template.In order to establish an accurate template,a way of human-computer interaction is proposed to correct,which can effectively solute the impact of light,noise,etc.Then unplaced screws image (template image)subtracts the detected image,standing out the features of placed screws.Finally,template matching is used to position the screws.Experimental results show that this method can effectively solve the impact of shelves damaged,screws oxidation and noise interference,and realizes simple.Counting accuracy rate of this method can achieve above 98%,and average detection time is 3.67 ms,significantly improving warehouse management capabilities.
参考文献
| [1] | 张少军;苟中魁;李庆利;李忠富;金剑.利用数字图像处理技术测量直齿圆柱齿轮几何尺寸[J].光学精密工程,2004(6):619-625. |
| [2] | 侯维岩;张利伟;党蟒;段政伟;宋杨;王海宽.一种基于图像处理的棒材计数测量系统的设计与实现[J].仪器仪表学报,2013(5):1100-1106. |
| [3] | 郭杰;雷刚;陈健生;向守兵.基于机器视觉的钢包头检测系统设计[J].液晶与显示,2013(5):770-775. |
| [4] | 段瑞玲;李庆祥;李玉和.图像边缘检测方法研究综述[J].光学技术,2005(3):415-419. |
| [5] | 张达;谢植;艾江山.棒材在线计数中断面定位方法研究[J].仪器仪表学报,2010(5):1173-1178. |
| [6] | 周群;胡光锐.基于模板覆盖法的钢筋在线计数方法[J].上海交通大学学报,1999(11):1462-1464. |
| [7] | 林玉池;崔彦平;黄银国.复杂背景下边缘提取与目标识别方法研究[J].光学 精密工程,2006(3):509-514. |
| [8] | 孙可,张钧萍.局部自适应方向模板匹配的高分辨率遥感影像道路提取[C].2015年光学精密工程论坛论文集,2015:509-515. |
- 下载量()
- 访问量()
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%