少有人走的路这篇是vidi的宣传文档，勇哥引用至些，方便各位观看。
目前在网上，vidi的和谐版是机密，大量骗子打着销售此软件的名号骗钱，这从一个侧面证明这款软件的效果上的成功。
各路神仙都从这套视觉软件上面看到了商机，是因为实它在视觉处理方面效果是前无古人的。
使用vidi如同“像人脑一样识别”！
A revolution is now underway and why is it possible only now? It is thanks to the presence of those 3 elements simultaneously (Abundance of Data, Affordable computing power and deep learning theory)
Many companies have invested / are still investing in this field in order to develop product (for some we are using them almost every day)
Siri as a voice recognition
Google Sef-driving car
IBM Watson who won in 2011 the game Jeopardy
And many more such as … (automatic labeling of photos by Apple and Facebook) and also
ViDi Suite for the industrial World …
A place crowded with people, many of which are visual quality inspectors.
So, why are there so many inspectors with their pink jackets, standing next to magnifying glass lamps?
Because, contrary to what we think, there are many visual inspection tasks that are difficult, if not impossible, to automate today.
Bat that is going to change!
Our Software as A unique architecture and This image processing software is composed / has an architure composed by a core technology which can be use through different ways/ with different focuses:
1.To find an element/characteristics/feature within the image (what it is and where it is?) 2.To identify a region of an image (what is different from the norm ?) 3.To describe/analyse the image itself (what it is?)
ViDi blue:
ViDi red:
ViDi green:
sorting of beer bottles according to vendor identity (recycling)
OCR(Optical Character Recognition)
识别率 > 98%
(with error correction)
OCR 案例
Google 街拍门牌号
钢铁的跟踪
零件的定位
低照度下的定位
有遮挡物的定位
高反光材料的定位
工件定位
§像人一样定位
§通过自学习的方式，无需编程，仅仅点击选择就可定位
§ 功能强大，解决了复杂环境下的跟踪和识别
