Attention

You are viewing an older version of the documentation. The latest version is v3.3.

Real Time Vision

The real-time vision provides a deterministic way to complete synchronization between motion control and image capture even the objective is in high-speed moving. Then the stopping time can be saved, thus to improve efficiency and productivity.

The key to achieve the synchronization is the time-aware IO. And an EtherCAT IO with timestamping can be utilized to trigger a deterministic capturing for an accurate image. Then the accurate image can be processed by machine vision to provide precise position offset and angle offset for next-step motion control.

Usage case

../_images/usage.png

In the SMT production line, this kind of operation is always shown, that the gantry with a sucker sucks a chip from the plate and then mounts on the PCB. But it’s not always able to hit the expected point and the expected angle of the chip during the suction. Even little shift can lead to deviation, so that the chip cannot be mounted in the right place later. Then the machine vision does help after the image capturing, to computer position/angel offset value of the chip for perfect mounting.

In the tradition way, the gantry will stop above the camera and wait for a while for image capturing. Now is not necessary anymore when applying real-time vision. And the efficiency can be improved significantly.

Work Flow

The application controls the motion by EtherCAT and synchronizes with IO. When reaching the target timing, IO will trigger the camera to capture an image. And then the image will be processed with machine vision to provide the position value and the angel value. By data exchanging, the application continues the motion control with position information.

../_images/workflow.png

Solution principle

../_images/trigger_timing.png
../_images/moving.png

  • T0 is the time when CPU sends shooting command

  • S0 is the position to prepare trigger shooting, which can be read in cyclic task. It corresponds to T0.

  • T1 is the time to trigger the picture shooting

  • S1 is the expected fixed position to trigger shooting.

  • △T can be calculated with ∆𝑇=(𝑆_1−𝑆_0)/𝑉

  • T2 is the time when the image is captured on CMOS

  • S2 reflects the real position where the image is capturing.

  • The time between T2 and T1 is used for camera exposure and image generating

  • Real-time vision should make the time intervals (T2 - T1) and (T1 – T0) deterministic

Demo example

The demo code is integrated into IgH EtherCAT stack components as an example in examples/fly_trigger_poc