x11vnc 边界框坐标

Question

I have a question regarding screen regions or possibly mouse/cursor coordinates in x11vnc.

I am trying to create all possible mouse positions using an image within an image. The primary image will always be 765 by 503 pixels. The secondary image is unknown until I figure out how to extract the secondary region in a easily-reproducible manner. Meaning, I will need some way of somewhat some-what accurately grab the secondary image that I want, and then be able to extract the x,y of the secondary image in respect of the primary image. Meaning that the bottom-left corner of the secondary image coordinates of x,y will not be 0,0 but rather the x,y values as if it were the main image. This is only meant to help create accurate coordinates, actual results may differ.

I know the image will be there because I am using vncdotool to expect the image, and then perform an operation once the image is found. Note: I am not sure what what rexpect within the vncdotool does

Using x11vnc I have shared a single application, so all coordinates are needed, I just need to figure out a way to map image(s) to coordinates upon the vncdotool expect finding the secondary image

Answer 1

This github repo solves this question

You can create an XML file using a simple gui on the image to draw the rectangle you wish to create the bounding box on and get the x_min, y_min, x_max, y_max pixel coordinates which can translate into mouse/cursor coordinates.

https://github.com/tzutalin/labelImg