锅仔片表面指纹检测方法

Metal Dome Surface Fingerprint Detecting Methods 

摘要:

锅仔片是薄膜开关、微型开关等电子元件中的关键部件之一,具有接触平稳、导通性强、回弹稳定、手感俱佳的几大优点；在锅仔片的生产过程中,不可避免地会被人的手指接触到,并留下指纹和油脂,导致锅载片加快氧化,影响其寿命,对主体产品的质量留下了隐患；因此,锅仔片的质量检测是提高主体产品竞争力的重要环节；目前,生产企业大多采用的检测法是人工目视检测法,该方法效率低,且容易误判；为了提高生产效率,提出了基于机器视觉的锅仔片表面指纹检测方法；该方法通过利用广义霍夫变换对目标进行定位、提取目标的灰度共生矩阵和支持向量机分类识别,实现了锅仔片质量的自动检测,并给出了该方法的工作流程；实验结果证明,该方法取得了良好的效果,具有一定的实用价值和推广价值。

Abstract:

Metal dome is one of the key components of electronic components,like membrane switch,microswitch,etc.It has several big advantages,for example,smooth contact,strong connectivity,stable spring-back and delicate touch .In the process of metal dome production,it’s inevitable to be touched by the finger of people . As a result, fingerprints and grease are left on it,leading to accelerated oxidation to metal dome,short working life,and other potential damage.Therefore,metal dome quality testing is very important to improve the competitive power of the products.At present, most of manufacturing enterprises use the human visual test,which has low efficiency,and is easy to be wrong.In order to improve the production efficiency,a metal dome surface fingerprint testing method based on machine vision proposed.The method by using the generalized Hough transform locates the target position,then extract the gray level co-occurrence matrix and supporting vector machine classification,which realizes the metal dome quality automatic test,and the working flow of this method is given.The experimental results show that this method has obtained a good effect,possessing certain practical value and popularization value.