ng 
all 
ge 
ns 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XL-5/W2, 2013 
XXIV International CIPA Symposium, 2 — 6 September 2013, Strasbourg, France 
Detected areas appear so in the mask as white areas. As all 
positions are tested in a XYZ file, the mask has the same 
dimension as the XYZ file and so as the attached image (Figure 
13). 
(a) (b) 
  
  
  
Figure 13: Image (left) and associated mask by the propagation 
of annotation (right) 
In the event that the drawn area (corresponding to the 
annotation) does not appears in one of the other images, the 
created mask for this image will be only composed of black 
pixels. 
The annotated area can be displayed by superimposing the 
image and the mask, and by affecting a transparency value on 
the mask and a color value on white pixels (Figure 14). 
Figure 14: Visualisation of the area on another image 
In this process, the transfer can be performed in two ways for an 
image: from the picture (definition of an annotation) or to the 
picture (transfer from another picture). 
5.4 Multi-view enrichment of annotations 
This transfer only allows the search of existing points on the 
annotated image. But the object to be annotated frequently does 
not appear wholly in any image. Indeed, in these cases, other 
views are needed to completely select this object. 
For this reason, the objective of multi-view enrichment is to 
permit the user to define an annotation from different views 
while using the propagation's method previously presented. 
A first list of X, Y and Z coordinates is extracted by 
implementing the steps 5.1 and 5.2 on this view. Then the steps 
5.1 and 5.2 are implemented again on another view and a 
second list of X, Y and Z coordinates is extracted. By grouping 
together these two lists, a third one, representing the whole 
annotation, is created. Finally the step 5.3 is processed for all 
images by using the combination of the two extracted lists 
(Figure 15). 
419 
  
  
2 m : x 
» p 
e od 
Figure 15: Steps of multi-view definition: (a) definition of a part 
of the annotation on a view and search of X, Y and Z 
coordinates (pink), (b) definition of a second part of the 
annotation on another view and search of X, Y and Z 
coordinates (blue), (c) association of the two list of coordinates 
(green), (d) projection of the modified annotation on others 
images 
These steps can be generalised for the use of more than two 
images. It is sufficient to repeat steps 5.1 and 5.2 on each 
concerned images and, then, to assemble all the lists of 
coordinates before implementing step 5.3. 
Thus, this enrichment of annotations can be performed from 
different viewpoints and the definition of an annotation can be 
defined at best. 
6. CONCLUSION AND PERSPECTIVES 
This work has shown a process based on 3D information to 
transfer annotation towards a set of spatially oriented pictures of 
a building. Despite the results obtained with this study, some 
issues need to be resolved and some reflections should prompt 
further research. 
First of all, in order to improve the definition and the transfer of 
an annotation, an automatic segmentation of the image or of the 
implicit point cloud (implicit because contained in the XYZ 
files) could be envisaged. For example, with an efficient 
segmentation of images, the definition of annotations could be 
implemented by selecting parts of the segmentation and the 
transfer could be implemented by detecting parts of the 
segmentation instead of detecting pixels. 
Besides, as different levels of semantic description could exist, 
the overlapping of annotations should be envisaged and 
managed. 
Then, a set of 2D or 3D analyse tools (color, shape ...) could be 
developed, with the help of images or point cloud. 
Afterward, the adding of new photographs to the already 
annotated images should be provided by the system. 
Furthermore, as the state of a building evolves in time, the 
management of images from different time can be expected. 
At last, if annotations are semantically defined, several queries 
(by single annotation, by terms ...) could be formulated with the 
crossing of all data according to different criteria.