Perspective Match no dialog box as per video

how does one get to appear the dialog box we see in the shipping crate video demo, where the done button is greyed out until her fifth pick, she doesnt make mention of the box and how to acquire it.

Could someone redo this with a real life subject, say a truck with cab and no engine housing, just chassis with front and back wheels as a model and show how to match it to an image of the completed truck. cant use front wheels as they can vary in angle. Need the model to match the image so as to decide where the features sit on the model and finish its creation. Easier to listen to as well, as this is bad audio.

Shipping crate video is effectively a finished model matched to the target picture.


Just buy the Mac version of Rhino 5 :wink:

How does one do ‘done’ on windows ?


Hi Steve,

I suppose that is the possible after the 6th point picked by hitting Enter:


I was with 5 points and not able to find a 6th, guess I shall have to retry, wish it said 6 points need to be found !

I still would like to see this command shown working or shown up !
Could someone redo this with a real life subject. Not a box.


FWIW, I cannot get this command to do anything useful in SR13. I tried in the WIP 3 days ago and that went fine but trying now with the same picture is not working at all.

Steve, still not understood?
PM always expected a BBox ! Namely to place reference points an drawing on there faces.
PM is not a magic tool and not 100% exact (camera position, distance and angle unknown).
The BBox can not have any dimensions than the same from intersections from Two Point Perspective.
If your 3D object not have a BBox with correct dimensions, matched to intersections in your picture, then nothing will match.

SketchUp has advanced PM control, even with zoom, automatic mapping from PM to faces and can be exported to Rhino.

My recollection is the math of determining the location, orientation, and lens angle of view requires a minimum of six points on the photo be matched to point locations in 3D space. If straight lines can be matched the number of points can be reduced. That’s a basic requirement, not something smarter or more clever software can relax.

Hi, Eddi,

I mentioned somewhere an aircraft, take a spitfire for example, cant box that. If I have it partly drawn, say wing, fuselage, no engine no tailplanes, how do I establish a box on it ? I need to select points on the picture and on what is created so far and have one matched to the other, so as to work out length and look of engine, tailplanes etc.

I also have an old truck, cab but not engine, chassis, rear and front wheels which are different track sizes, with little to go on apart from this to match to the photo, but needing to match to photo to create rest of it, …thats the task.

The bus is again an easy subject, a box. But life isnt so easy.

I cant find out if free sketchup has the PhotoMatch and one cant ask question until having bought sketchup pro. cant afford $100s for matching a photo.


PerspectiveMatch in Rhino requires at least six points in the photos to be matched to already existing points in Rhino. My understanding is these points do not have to corners of a box. If you don’t know already know the relative locations of at least six points then you can’t use PerspectiveMatch.

PhotoModeler software works with multiple photos without knowing anything about the geometry in advance, and also some single photos of box like objects with suitable perspective.

Now using 6 points on my truck cab, it displays the ability to hit enter, however results in a ‘shambles’.
yet points were at roof corners, cab base corners, good locations of an accurate model and the pic is accurate.


Do you already have the 6 points accurately modeled in Rhino, or are you try to match to a simple box?

Also, do you know for certain that the photo has not been cropped?

The 6 points are carefully placed on what is a very accurate recreation by me of the cab, based on site measurements of the real thing also by me. They are at locations that go with perspective vanishing lines, e.g either corner of cab frontage at top. The photo is out of museum archives and is a scan of an original large quality print, given the detail in the image its not a scan of part of the print, as it is the subject matter for the photo. If they didnt quite capture the last 1mm or so of the print edge that wouldnt make any difference I reckon.

I have however also some other photos which are crops, as the vehicle was to one side of the main subject, and paying double for the raster file when I needed only the vehicle was non sensical, As such they wouldnt work, but if I were to have the entire image, they wouldnt be big enough to work on, trouble with this method is no ability to zoom in on the photo to carefully place a ref point or make use of the detail and be accurate, then when done, to plot the required detail onto the model.