AUTOGRAF—AUTomated Orthorectification of GRAFfiti Photos

Round 1

Reviewer 1 Report

The authors discuss the application of AUTOGRAF, an automated and freely-available orthorectification tool which converts conventional graffiti photos into high-resolution, distortion-free, and georeferenced graffiti orthophotomaps, developed in the framework of INDIGO, a research project on the graffiti.

Their work is interesting, in particular for further potential applications on cultural heritage, when handling many images, that need corrections for various distortions caused by perspective and surface topography. The diffusion of the proposed approach is surely facilitated by the public freely availability of the source code.

For these reasons I consider their work suitable for publication on Heritage.

Remarks:

the "Materials and Methods" section is quite long and it is a bit dispersive for the reader. For this reason I've have evaluated with "can be improved" some questions proposed by the default peer-review survey. Nevertheless, the paper can be published in the present form. However, if the other referees will require some revisions, I suggest to the authors to take this opportunity to try  improving the ease of reading of this section.  

Author Response

Dear reviewer, 

Thank you very much for reviewing our manuscript. Please see the attachment for our responses. 

All the best, 

Benjamin Wild on behalf of all co-authors

Author Response File: Author Response.pdf

Reviewer 2 Report

In this MS the AUs describe software convert conventional graffiti photos into high-resolution, distortion-free, and georeferenced graffiti orthophotomaps.  The software is referred to as the acronym AUTOGRAF.  AUTOGRAF is written in Python and is built as an add-on to Agisoft's Metashape Professional commercial 3D photogrammetric software (Angisoft LLC, St. Petersburg, Russia).  The AUTOGRAF code can be freely downloaded for use.  The MS is well written without the need for English language assistance.  I believe it will be of interest to the readers of the special issue.  I offer the following comments on the MS.

L41       It is interesting that the AUs distinguish between people and politicians.

L46       Graffito is defined as writing or drawings scribbled, scratched, or sprayed illicitly on a wall or other surface in a public place.  To classify prehistoric cave paintings as graffiti (illicit) is unfounded.

L141     I am confused by this statement.  I thought placing the grafitti in context of its surroundings was important? It is quite a subjective statement to assume certain large objects do not add any value, just to save disk volume.  I propose that a reason why grafitti is found in one location over another is not solely the presence of a flat surface, but to provide privacy during the creation.  

L456     I would have chosen a lossless image format instead of the minimally lossy jpeg. 

 General: Graffiti is also left on irregular shaped (non-planar) surfaces such as statues.  In this case, a graffito is making an additional statement against the statue.  Can AUTOGRAF be applied to these irregular surfaces?

Author Response

Dear reviewer, 

Thank you very much for reviewing our manuscript. Please see the attachment for our responses. 

All the best, 

Benjamin Wild on behalf of all co-authors

Author Response File: Author Response.pdf

Reviewer 3 Report

The authors investigated a novel method for documenting graffiti art using photogrammetry-based techniques. The study is well investigated and the paper is well drafted. The topic indeed brings interest to the field. The reviewer has a few comments that may be helpful for improving the quality of the paper:

1) There are a few related works in the domain of photogrammetry-based documentation of graffiti art in the literature. Could the authors review and differentiate them in the paper? This will strengthen the merit of the proposed method.

2) Is INDIGO a name of a research project? This shall be clarified with some explanation of the background.

3) Does AUTOGRAF represent a self-developed data processing platform, an integration of algorithms, a software add-on feature, or something else? After reading Section 2.1, the reviewer has the impression that most of the work is done in Metashape, while a small portion of the tasks is performed via AUTOGRAF (perhaps Section 2.1.5). How exactly this AUTOGRAF work? What automated feature exists in this method? What efforts are required to implement AUTOGRAF? Shall the point cloud be re-loaded to a different platform for this AUTOGRAF process? Did the author develop the algorithms in AUTOGRAF or it is built upon some existing tools? The review likes to see  details of the proposed AUTOGRAF as it relates with the merit of this study. Also, it would be wise to differentiate work in AUTOGRAF vs work in Metashape in some regards as Metashape is the off-the-shelf engineering tool.

4)  The reviewer would propose to remove Option B and C from Fig.2 as they are something that has never been investigated in this paper. These ideas can be discussed in the later part of the paper (which the authors already did).

5) How do the authors obtain RMSE in Fig. 7? Is this from Metashape or somewhere else?

6) Section 2.1.5. Why do the authors decide to filter out points away from the wall (e.g., isolated structures)? The reviewer did not see a discussion on the motivation. Is this because these features may not be accurately presented in the orthophoto or try to reduce the amount of work?

7) When applying this method to other cases as discussed by the end of the paper, does it limit to wall-associated heritage objects? For a complex historic site such as a building, bridge, or an outdoor environment, does the proposed method has challenges to create an accurate orthophoto of large scope, non-plane-like scene with complex 3D configuration? 

Author Response

Dear reviewer, 

Thank you very much for reviewing our manuscript. Please see the attachment for our responses. 

All the best, 

Benjamin Wild on behalf of all co-authors

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

I recommend this paper for publication.