首页 >> 技术支持 >> 机器视觉检测知识普及

阿里巴巴

联系电话


相关推荐:


五金行业视觉检测


电子行业视觉检测


LED行业视觉检测


注塑行业视觉检测


包装行业视觉检测


视频案例:







机器视觉检测面临的十大挑战(二)

时间:2018-03-13  来源:本文搜集于网络,由机器视觉检测设备厂家创盈时代整理,欢迎转载,转载请注明出  点击:207

上一节我们讲过机器视觉检测面临的十大挑战上半部分,相信很多小伙伴已经有过初步了解,今天我们继续来说说下一节,更深入的解剖视觉检测在目前面临的一些问题。
In the last section, we talked about the top ten challenges of machine vision inspection. I believe a lot of small partners have already had a preliminary understanding. Today we continue to talk about the next section, and further dissect some problems faced by visual inspection.

闭塞:遮挡意味着物体的一部分被遮住了。在前面的四个挑战中,整个对象出现在相机图像中。遮挡是不同的,因为部分对象丢失。视觉系统显然不能检测到图像中不存在的东西。
Block: occlusion means that a part of the object is covered. In the four previous challenges, the whole object appears in the camera image. Occlusion is different because part of the object is lost. The visual system obviously can't detect something that does not exist in the image.

有各种各样的东西可能会导致遮挡,包括:其他物体,机器人的部分或相机的不良位置。克服遮挡的方法通常涉及将对象的可见部分与其已知模型进行匹配,并假定对象的隐藏部分存在。

There are all kinds of things that may cause occlusion, including the other objects, the parts of the robot, or the bad position of the camera. The method of overcoming occlusion usually involves matching the visible part of an object with its known model, and assuming that the hidden part of the object exists.


CCD机器视觉检测相机


比例:在某些情况下,人眼很容易被尺度上的差异所欺骗。机器视觉检测系统也可能被他们弄糊涂了。想象一下,你有两个完全相同的物体,只是一个比另一个大。想象一下,您正在使用固定的2D视觉设置,物体的大小决定了它与机器人的距离。如果您训练系统识别较小的物体,则会错误地检测到两个物体是相同的,并且较大的物体更接近相机。
Proportions: in some cases, the human eye is easily deceived by the difference in the scale. Machine vision testing systems may also be confused by them. Imagine that you have two exactly the same objects, only one larger than the other. Imagine that you are using a fixed 2D vision setting, and the size of the object determines its distance from the robot. If your training system identifies a smaller object, it is wrong to detect that two objects are the same, and the larger objects are closer to the camera.

尺度的另一个问题,也许不那么明显,就是像素值的问题。如果将机器人相机放置得很远,则图像中的对象将由较少的像素表示。当有更多的像素代表对象时,图像处理算法会更好地工作,但有一些例外。
The other problem of the scale, perhaps not so obvious, is the problem of pixel value. If the robot camera is placed far away, the object in the image will be represented by a small number of pixels. When there are more pixels to represent objects, the image processing algorithm works better, but there are some exceptions.

照相机放置:不正确的相机位置可能会导致以前出现过的任何问题,所以重要的是要正确使用它。尝试将照相机放置在光线充足的区域,以便在没有变形的情况下尽可能清楚地看到物体,尽可能靠近物体而不会造成遮挡。照相机和观看表面之间不应有干扰的背景或其他物体。
Camera placement: the incorrect position of the camera may lead to any problems that have occurred before, so it is important to use it correctly. Try to place the camera in a well lit area so that objects can be seen as clearly as possible without deformation, as close as possible to objects without causing occlusion. The background or other objects that have no interference between the camera and the surface.

运动:移动有时会导致计算机视觉设置出现问题,特别是在图像中出现模糊时。例如,这可能发生在快速移动的传送带上的物体上。数字成像传感器在短时间内捕获图像,但不会瞬间捕获整个图像。如果一个物体在捕捉过程中移动太快,将导致图像模糊。我们的眼睛可能不会注意到视频中的模糊,但算法会。当有清晰的静态图像时,机器视觉检测效果最佳。
Movement: movement sometimes causes problems in computer vision settings, especially when blurred images appear. For example, this may occur on a fast moving object on a conveyor belt. The digital imaging sensor captures the image in a short time, but does not capture the entire image in a moment. If an object moves too fast during the capture process, the image will be blurred. Our eyes may not notice the blur in the video, but the algorithm will. When there is a clear static image, the effect of machine vision detection is the best.

期望:与视觉算法的技术方面相比,最后的两个挑战更多地涉及到您的视觉设置方法。机器人视野面临的最大挑战之一就是工作人员对于视觉系统能提供什么不切实际的期望。通过确保期望符合技术的能力,您将从技术中获得最大收益。您可以通过确保员工接受关于视觉系统的教育来实现这一点。
Expectation: the last two challenges are more related to your visual settings than the technical aspects of visual algorithms. One of the biggest challenges facing the visual field of robots is the unrealistic expectations of the staff of the visual system. By ensuring the ability to meet the technology, you will get the greatest benefit from technology. You can achieve this by ensuring that employees receive education on the visual system.

深圳市创盈时代科技有限公司是一家专注于机器视觉,视觉检测设备,CCD光学检测设备,机器视觉检测,视觉方案定制,非标自动化设备的研发生产和销售于一体的国家高新技术企业。

原文地址:http://www.chytime-ccd.com/zhishi/287.html

本文关键字:机器视觉检测

Copyright @ 2016 版权所有:深圳市创盈时代科技有限公司 粤ICP备13087279号-4
联系人:汪小姐 电话:0755-81779389 手机:18819037886
地址:深圳市沙井沙一社区万安路长兴科技园16栋2楼 网址:http://www.chytime-ccd.com
主营产品:CCD视觉检测  机器视觉  视觉检测系统  CCD视觉检测设备  影像检测系统  网站地图  XML