Tell us your PC-DMIS Vision Enhancement Suggestions

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Tell us your PC-DMIS Vision Enhancement Suggestions

    Although a similar thread exists on the main PC-DMIS forum, I thought it would be useful to get some vision specific feedback.

    Internally, we have around 200 ideas for Vision improvements, some radical and big, some smaller but still of benefit. In order to help prioritize/schedule them, it would be good to find out which are perceived to be the most important to the most people.

    Some questions to get you thinking :-

    What can't you do that you want?
    What would make programming easier?
    What Auto Features would you like available for a Video Probe?
    What wonderful new enhancement idea would you like to see?
    How important is blob analysis?
    How important is OCR?
    More filtering needed?
    Could Screen Captures be improved?
    Are diagnostics tools sufficient?
    Is AutoTune useful and how could it be improved?
    Should we try to optimize measurement speed?
    Do we give too much control over focus measurement (i.e. do we make it too easy to get the wrong result by going too fast)?

    Thanks in advance for your input.

  • #2
    3 Or 4 Points Instead Of A Whole Diameter

    It would be nice to click multiple points instead of going through the whole circumference of the diameter when starting a program.
    When you have a 2" diameter it takes long to go through the diameter a portion at a time.

    Comment


    • #3
      Hi JacobMag, I'm not 100% sure I understand the question.

      Are you talking about clicking on the live view when programming your feature (which is then measured with Automatic Target (auto edge detection))? You can click anywhere on the circle when defining it, although the more spaced out the "hits" are, the more accurate our estimation of the circle will be. You can also double click on the circle in the live view, and it will define a target for the part of the circle in the Field Of View. You can then edit the start and end angles in the auto feature dialog to turn it from an arc section into a full circle.

      Or are you talking about measuring a circle in Manual Target (crosshair) mode? If so, when you program the feature, you can drag the position of the targets around the feature to be where you want them to be measured at.

      Or, are you saying that at execution time, it takes too long to go around the big circle taking lots of hits automatically? If that's the case you can try a few things :-

      Measure the feature at a lower magnification
      Ensure your target bands are only as wide around the feature as neccessary to ensure you find the correct edge. If they're bigger, it will take more processing time.
      Use Specified Edge detection method if the edge is not noisy (it's quicker than Matching Edge).
      Split the donut target into sections (right click on it and say Insert Target). Then, right click on the target and set it to Point Density "None". Do this as many times as you need to section off different bits of the circle you're not interested in. In a future version, we plan to make it easier to do this.

      If I've missed the point of the question, please give more information.

      Thanks for your input!

      Comment


      • #4
        Automatically distribute multiple targets accross a circle

        Hi

        We are just starting to use PC-DMIS.

        During our learning, we had a part which has an inner circle with cuts in it Similar to a gear.

        To measure it, we have to define a auto circle and then divide it in a big number of targets, twice as much as the gear has tooths, to only measure the sections, where the circle actually exists.

        It would be great if we could specify the first section and then tell PC-DMIS in what angle steps it has to repeat this section.

        This could be done using some kind of an assistant to create the targets.

        An other solution would be to measure a circle segment in a loop and then use the circle feature array to compute the circle (maybe this is possible already, if so, please explain how to do it?).

        Also I was looking to find the target specification in the PC-DMIS command mode and couldn't see where the definitions for the targets are.

        Regards,

        Herbert

        Comment


        • #5
          Hi Herbert,

          We are planning an enhancement for V4.3 that would be perfect for the gear section problem.

          Another way to do it currently, would be to use the paste with offset. Create a circle that measures one of the teeth. Then, define a pattern that rotates around the centre of the circle at the specified angle increments. Then copy the feature and select paste with pattern.

          It depends on which version you're using as to where the target values are. Also note that some parameters (specifically the relative position of the targets around the nominal form) are not shown in the editor, however the parameters for how to measure each of these targets (illumination, edge detection params, etc) are in there. In V4.2 you need to expand the SHOW_VISION_PARAMETERS section. There is a similar term to expand in V41 and prior versions.

          Comment


          • #6
            Hi

            Hi

            We are using PC-DMIS 4.1.

            When will 4.3 be available?

            After I do the paste with offset, how do I construct my final circle from the various sectors?

            Regards,

            Herbert

            Comment


            • #7
              A couple things I would like to see changed in the next version are crosshairs instead of a pointer. As a dedicated off-line user and having some 10 plus yeares doing this sort of thing I've found the software that has Crosshairs make it infinately easier to tell where that touch point center is going to be.
              Second, I'd also like to be able to move stuff using transform by level. Restrict the Transform to only the active levels. In short, if a level is turned off, transform doesn't touch it.
              And the third Item I'd like to see is the ability to assign my mouse buttons. I don't use a 3 button mouse because my job requires me to use other cad software as well. I use an 8 button trackball and I've found that PCDmis buttons are exactly backward to most of the other CAD software I use and it it the only one where button reassignment is not possible. This gets frustrating when going between applications. Muscle memory is real. I'd like to be able to reassign the buttons.

              Comment


              • #8
                The new CMM-Ve has different offset than CMM-V6, and doesn't exist in the probe.dat.
                Can you post the new offset ? Is there a new way to calibrate angles, or do I use the same star tip ?
                I need those informations to auto calibrate probe, and use it with a tesa-star_m.
                Thanks !

                Comment


                • #9
                  CMM-Ve is present in probe.dat of 2010MR1 !

                  Comment


                  • #10
                    Originally posted by nkayWAI View Post
                    Although a similar thread exists on the main PC-DMIS forum, I thought it would be useful to get some vision specific feedback.

                    Internally, we have around 200 ideas for Vision improvements, some radical and big, some smaller but still of benefit. In order to help prioritize/schedule them, it would be good to find out which are perceived to be the most important to the most people.

                    Some questions to get you thinking :-

                    What can't you do that you want?
                    What would make programming easier?
                    What Auto Features would you like available for a Video Probe?
                    What wonderful new enhancement idea would you like to see?
                    How important is blob analysis?
                    How important is OCR?
                    More filtering needed?
                    Could Screen Captures be improved?
                    Are diagnostics tools sufficient?
                    Is AutoTune useful and how could it be improved?
                    Should we try to optimize measurement speed?
                    Do we give too much control over focus measurement (i.e. do we make it too easy to get the wrong result by going too fast)?

                    Thanks in advance for your input.
                    Really? I'd love to be able to suggest enhancements for PC-DMIS (CMM)... especially if someone from WAI really reads and respond to them.
                    PC-DMIS CAD++ 2o19 R1 SP11

                    Comment

                    Related Topics

                    Collapse

                    Working...
                    X