> For the complete documentation index, see [llms.txt](https://textav.gitbook.io/textav-event/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://textav.gitbook.io/textav-event/unconference-projects/ai-pipeline/i-learned-what-tesseract-can-do-and-so-can-you.md). # I learned what Tesseract can do (and so can you!) {% embed url="" %} ## I learned what Tesseract can do (and so can you!) Made at the Text AV conference at NYU July 19-21 Corbin Ordel
Programs will use: [VDXM5](http://vidvox.net/), [OCR-OSC](http://www.davidlubl.in/projects/#/ocr-osc/), [OSC-WEB](https://github.com/automata/osc-web)
Using a couple programs that are downloadable and easy to use (they have desktop icons and everything), you can create your own OCR (text recognition) program!
Step 1: Download and Install this program called [VDXM5](http://vidvox.net/), (Or any program that can send video through a syphon as [OCR](http://opensoundcontrol.org/introduction-osc) communication) Step 2: Download and Install this program called [OCR-OSC](http://www.davidlubl.in/projects/#/ocr-osc/). These programs were created by a very cool person named [David Lublin](http://www.davidlubl.in/). He has made some great projects including this [one](https://twitter.com/TVCommentBot/)! Step 3\*\*\*: Go to and follow the directions for install and usage: for your enjoyment and terminal ease - here is a picture of what to do (if you have git and npm package manager installed on your machine)! ![](https://lh5.googleusercontent.com/TcHJwsWZEKMpUqvKUiOxm8q9fYJoyrdVJGUsQOJHHTxytMhLLUrt006YoHrqvjWp1GLtus27H8n2DyF6gJBD66GmECb59K4-K3Zwt_aUMowXV83rJeV1dWUNnYgpbivw-31m93G3) (\*\*\*This step is optional, but if you want to see your collected data in your terminal, do this! Also why not do it anyway! it doesn’t take long.) \~And I quote, “The objective of osc-web is to make possible to send and receive OSC messages on the Web browser.”\~ This is very useful for all sorts of cool things**.**\
Step(‘s) 4 (Set up of all components): NOTE: If you want to just watch a video walkthrough, click here ---> {% embed url="" %} * Open VDXM5 and make a new project (file --> new) * Drag your video into your media bin * Go to “Plugin” and select “Syphon Output” and select “Main OutPut” * Click “Layers” and find the “Crop” button - use the cropping feature to find your desired text, in the case of the attached demo, it is the ticker at the bottom of the hyper relevant cnn video. * After your video is cropped, Open OCR-OSC * Go back to VDXM5 and go to file → preferences → OSC * Click out “Output Ports” and make sure you have something that looks like this: ![](https://lh5.googleusercontent.com/QmTDypwqHXWdDuc4KoPVWuFraOK-hLgpKaS6MsErhkQ4Zagvuc0qowRZxfWnEqPlLW60U6GYOpRxwOolpMmsJXm3RoAuQ2JJCXZyFOdBk9bt513_A7sJmsxMaFyE61hz3wwyD6BR) * Go back to OCR-OSC and click the drop down menu - you should be able to select “VDMX5 - Main Output” ![](https://lh3.googleusercontent.com/RoP4hAEzLc92ZYSp-N8vtg6Vm94e4AwLJFXVUUdGZ65Jo3nRx21jMiU9tfpDCsGQFblc2ZlDUOEWQ43esMPUeENpQD9RY1r68pjiUIOQAhK2JtRwG4fI2VFmnpC7EovG7kZAD64g) * It should be streaming the video from VDMX5 when you select the correct source * Now what if we want information in our terminal? * Go to terminal and change your directory to osc-web (command is most likely \[ cd osc-web ] * But wait, no information? Where is it? * Go to you osc-web folder and find the web-side folder and click on app.html * But wait? Still no information? Go to preferences in OCR-OSC and change your port directed at 3333 * You should now be set up and watching your very own OCR (text recognition) program! Let me know if you need more info!!!
Thanks to all the awesome software that [@DavidLublin](https://twitter.com/DavidLublin) makes and the awesome people who came to [TextAV](https://sites.google.com/view/textav)! Email me @
### As google doc: {% embed url="" %}