Using GRIP For Vision Processing
GRIP is a GUI tool for developing computer vision algorithms. It's based on OpenCV, one of the most popular computer vision software libraries used for research, robotics, and vision algorithm implementations. With GRIP you will be able to execute nearly all the same algorithms that you would by just hand coding them in a text-based programming language.
The GRIP Interface
As I said before, GRIP is a GUI, this means you configure mostly everything without code. The GRIP user interface consists of 4 parts:
Image Sources are the ways of getting images into the GRIP pipeline. You can provide images through attached cameras or files. Sources are almost always the beginning of the image processing algorithm.
Operation Palette contains the image processing steps from the OpenCV library that you can chain together in the pipeline to form your algorithm. Clicking on an operation in the palette adds it to the end of the pipeline. You can then use the left and right arrows to move the operation within the pipeline.
Pipeline is the sequence of steps that make up the algorithm. Each step (operation) in the pipeline is connected to a previous step from the output of one step to an input to the next step. The data flows from generally from left to right through the connections that you create.
Image Preview are shows previews of the result of each step that has it’s preview button pressed. This makes it easy to debug algorithms by being able to preview the outputs of each intermediate step.
Generating Code from GRIP
After you have done what you wanted to do with GRIP, you will need to generate actual code for the algorithm you have constructed. To generate code, go to Tools > Generate Code.
Now you can create a vision pipeline and utilize the cool algorithm you made.
Notes and Tips
We used a Microsoft LifeCam HD-3000 Webcam
To connect to the robot's camera, add