In context capture can I just make a 2D map instead of making a 3D map, all I need is to get distances from objects?
You can produce an "orthophoto" that is a 2D orthographic view of your scene.
When producing an orthophoto, you can choose the sampling distance (i.e. the size of your pixels in your model units).
If your reconstruction is done at the right scale, then you will be able to easily make measurements.