Test automation

5 Great Automation Tools Based On Image Recognition

Another summary, this time to take to the workshop a set of tool for automated tests based on recognition of images. This type of treatment application corresponds to the type of Blackbox testing and allows you to quickly achieve the intended results.
I advise you to view all links included, their order is random, so read carefully. I have not seen a large operating systems based on the scripts in such technology. Whether they operate on a professional level? Are they reliable? What about mobile applications? What about changing the GUI, skins, themes? The questions are open…

Project Sikuli
Very simple script language connected with great screenshot and slice tool. Script is connected visualy with pictures. This a research project developed by people from MIT. Engine and IDE is written in Java so you can run it on every java based gui system. Includes great subcomponent / subpicture solution very helpful with checkboxes, radio buttons. I will make the review of Skiuli very soon, I am very impressed.

Well-known tool, mature and rapidly growing. You can read my article RoutineBot – Test Automation Tool Review – Technology does not matter?“. It allows very quickly create test scripts, and is reliable.

Enterprise class tool, very professional and mature. My article titled “Covers Everything? – Ranorex Automation Tool Review” shows only record and code level automation but it includes image based engine described in the following link: Image-Based Automation with Ranorex Recorder http://www.ranorex.com/support/user-guide-20/image-recognition.html

T-Plan Robot
Intresting tool formerly known as VNCRobot, based on VNC connection to test machine. GUI under test does not matter, VNC is available for nearly all software platforms – mobile too via PocketVNC. In addition to basic functions, the T-plan also includes the management of test cases. There are GPL and Enterprise version.

Another technology agnostic test automation tool. Again based on VNC connection. For now IDE is available for Linux and Mac, soon for Windows. Includes scheduler and image collection tool wchich enables you to collect, in example, icons with different statuses: disabled, selected, etc.


5 comments for “5 Great Automation Tools Based On Image Recognition”

  1. Nice Aggregation ..
    Question : Which do you most prefer. ?

    I personally have worked only with Ranorex and Sikuli ..
    I still like Ranorex it was more scalable, but wasn’t free.. 🙂

    Posted by Mithun Ashok | February 1, 2010, 3:28 pm
  2. Hi,

    Thanks for comment…

    The free tools usually lack of reporting and there is a problem with integration with TeamCity instance, CCNet etc.

    I’m not sure whether I can specify the preferred choice. It depends on many factors, to quickly perform some actions and select Sikuli or RoutineBot, if you want to arrange the whole system of quality assurance take Ranorex or T-plan.

    Posted by Marcin Zręda | February 1, 2010, 3:39 pm
  3. Hi,

    what do you think about EggPlant. It’s very expensive. Does it worth the price it asks for?

    what do you think about using image based automation tool for Flex/Flash applications?

    Posted by joe | February 18, 2010, 3:40 am
  4. Hi joe,

    Image based recognition is great for rapid test automation development but there are some limitations – as always. You cannot do everything, in example – test themed / skinned application. Some component are too complicated to test it with image.

    Eggplant is great tool this is complete engine, not only simple script with recognition component. I do not know how it supports your needs, try it and tell me …

    Posted by Marcin Zręda | February 18, 2010, 9:10 am
  5. Nice aggreation! Just to add another to the realm: iMacros includes image recognition for years. I have been using it since V5.x and I like it a lot: http://www.iopus.com/iMacros/irp/

    I use it for regression and performance tests of our Flash applets, and lately also for Silverlight.

    Posted by Mike Z. | February 28, 2010, 5:00 pm

Post a comment