Has anybody looked at adding support for a question type of "sketch" to ODK
Collect?
Perhaps something similar to the "barcode" type however this would a launch
a sketch app that would allow the user to draw a picture, add icons, take a
background picture etc. Presumably the result could be stored and uploaded
as a media type of image.
that is a great idea and along the lines of some things we are talking
about including being able to annotate a photo with arrows, boxes, and
text. Ideally, we will have the ability to have collect call other
phone apps to do specialized input types. If you don't mind, what is
your usage case? Just want to try to understand your
constraints/needs. For examples would you want to be able to edit the
sketch? Use geometric elements or just paint?
Thanks,
Gaetano
···
On Mon, Feb 14, 2011 at 4:36 PM, Neil Penman wrote:
> Has anybody looked at adding support for a question type of "sketch" to ODK
> Collect?
>
> Perhaps something similar to the "barcode" type however this would a launch
> a sketch app that would allow the user to draw a picture, add icons, take a
> background picture etc. Presumably the result could be stored and uploaded
> as a media type of image.
>
> Neil
>
> --
> Post: opendatakit@googlegroups.com
> Unsubscribe: opendatakit+unsubscribe@googlegroups.com
> Options: http://groups.google.com/group/opendatakit?hl=en
>
Perhaps all of those things. I haven't got a detailed set of requirements
for this, however the following uses come to mind:
Take a photo and place a generic marker somewhere on the image. The
question label/hint would provide context for the meaning of the marker.
Sketch a road intersection free style and add markers for POI's. These
markers could be specific to the sketch app and be transparent to ODK
collect.
Add a number of geometric shapes such as squares, lines, perhaps to
show houses and their position relative to a road. Colour code the houses
according to some schema described in the question label or hint.
The sketch capability would presumably be determined by the sketch
application, it may be possible to support multiple installed sketch
applications. Were you thinking of writing your own?
Ideally it should be possible to go back to the question and edit the
sketch. However that may be more complicated to implement then just having
the ability to replace the sketch.
Did you have a timeline for this work? I may be able to find a programmer
to help out if that suits.
regards
Neil
···
On Tue, Feb 15, 2011 at 11:48 AM, Gaetano Borriello < gaetano@cs.washington.edu> wrote:
Neil:
that is a great idea and along the lines of some things we are talking
about including being able to annotate a photo with arrows, boxes, and
text. Ideally, we will have the ability to have collect call other
phone apps to do specialized input types. If you don't mind, what is
your usage case? Just want to try to understand your
constraints/needs. For examples would you want to be able to edit the
sketch? Use geometric elements or just paint?
Has anybody looked at adding support for a question type of "sketch" to
ODK
Collect?
Perhaps something similar to the "barcode" type however this would a
launch
a sketch app that would allow the user to draw a picture, add icons, take
a
background picture etc. Presumably the result could be stored and
uploaded
as a media type of image.
Yes, I need a sketch answer type too. A bitmap is fine, but svg would be a
"nice to have"
Along with sketch, importing a static Google map image is another use case
for me. Here is the idea:
Using the GPS on the device, tap out to an external app (like how
barcode works today) but send along the lat/long as parameters.
Once on the map, use the pre-sent coordinates, or, provide simple
navigation tools (much simpler than current Google maps tools) to reset the
center point of the map. Zoom would be necessary too.
Once the location and zoom of the map are set right, "capture" the map
image and bring that back into the survey
Why do I need this if I have the lat/long already? The survey I am
working with needs to see the "plan" (ie overhead) view of a building (in
satellite mode) for buildings in the U.S. Once captures, the Collect user
should be able to paint on it as indicated in other comments in this thread.
Issues:
Google Static Map API key: depending on the size of the deployment,
you'd want each Collect device/deployment to use their own API key
Double-checking the terms of service for the Maps API
Whether this functionality is embedded in Collect using the Android
Maps library, or using the native browser and somehow easily saving the
image back into Collect, or developing an external app that does the work
and pushes the image back to Collect as the answer.
Stuart
···
On Mon, Feb 14, 2011 at 6:16 PM, Neil Penman wrote:
Hi Gaetano,
Sounds good!
Perhaps all of those things. I haven't got a detailed set of requirements
for this, however the following uses come to mind:
Take a photo and place a generic marker somewhere on the image. The
question label/hint would provide context for the meaning of the marker.
Sketch a road intersection free style and add markers for POI's.
These markers could be specific to the sketch app and be transparent to ODK
collect.
Add a number of geometric shapes such as squares, lines, perhaps to
show houses and their position relative to a road. Colour code the houses
according to some schema described in the question label or hint.
The sketch capability would presumably be determined by the sketch
application, it may be possible to support multiple installed sketch
applications. Were you thinking of writing your own?
Ideally it should be possible to go back to the question and edit the
sketch. However that may be more complicated to implement then just having
the ability to replace the sketch.
Did you have a timeline for this work? I may be able to find a programmer
to help out if that suits.
regards
Neil
On Tue, Feb 15, 2011 at 11:48 AM, Gaetano Borriello < gaetano@cs.washington.edu> wrote:
Neil:
that is a great idea and along the lines of some things we are talking
about including being able to annotate a photo with arrows, boxes, and
text. Ideally, we will have the ability to have collect call other
phone apps to do specialized input types. If you don't mind, what is
your usage case? Just want to try to understand your
constraints/needs. For examples would you want to be able to edit the
sketch? Use geometric elements or just paint?
Has anybody looked at adding support for a question type of "sketch" to
ODK
Collect?
Perhaps something similar to the "barcode" type however this would a
launch
a sketch app that would allow the user to draw a picture, add icons,
take a
background picture etc. Presumably the result could be stored and
uploaded
as a media type of image.
···
---------- Forwarded message ----------
From: Stuart Moffatt
Date: Tue, Feb 15, 2011 at 5:17 AM
Subject: Re: [ODK Community] Adding a "sketch" question type to ODK Collect
To: opendatakit@googlegroups.com
ODKers,
Yes, I need a sketch answer type too. A bitmap is fine, but svg would be a
"nice to have"
Along with sketch, importing a static Google map image is another use case
for me. Here is the idea:
Using the GPS on the device, tap out to an external app (like how
barcode works today) but send along the lat/long as parameters.
Once on the map, use the pre-sent coordinates, or, provide simple
navigation tools (much simpler than current Google maps tools) to reset the
center point of the map. Zoom would be necessary too.
Once the location and zoom of the map are set right, "capture" the map
image and bring that back into the survey
Why do I need this if I have the lat/long already? The survey I am
working with needs to see the "plan" (ie overhead) view of a building (in
satellite mode) for buildings in the U.S. Once captures, the Collect user
should be able to paint on it as indicated in other comments in this thread.
Issues:
Google Static Map API key: depending on the size of the deployment,
you'd want each Collect device/deployment to use their own API key
Double-checking the terms of service for the Maps API
Whether this functionality is embedded in Collect using the Android
Maps library, or using the native browser and somehow easily saving the
image back into Collect, or developing an external app that does the work
and pushes the image back to Collect as the answer.
Perhaps all of those things. I haven't got a detailed set of requirements
for this, however the following uses come to mind:
Take a photo and place a generic marker somewhere on the image. The
question label/hint would provide context for the meaning of the marker.
Sketch a road intersection free style and add markers for POI's.
These markers could be specific to the sketch app and be transparent to ODK
collect.
Add a number of geometric shapes such as squares, lines, perhaps to
show houses and their position relative to a road. Colour code the houses
according to some schema described in the question label or hint.
The sketch capability would presumably be determined by the sketch
application, it may be possible to support multiple installed sketch
applications. Were you thinking of writing your own?
Ideally it should be possible to go back to the question and edit the
sketch. However that may be more complicated to implement then just having
the ability to replace the sketch.
Did you have a timeline for this work? I may be able to find a programmer
to help out if that suits.
regards
Neil
On Tue, Feb 15, 2011 at 11:48 AM, Gaetano Borriello < gaetano@cs.washington.edu> wrote:
Neil:
that is a great idea and along the lines of some things we are talking
about including being able to annotate a photo with arrows, boxes, and
text. Ideally, we will have the ability to have collect call other
phone apps to do specialized input types. If you don't mind, what is
your usage case? Just want to try to understand your
constraints/needs. For examples would you want to be able to edit the
sketch? Use geometric elements or just paint?
Has anybody looked at adding support for a question type of "sketch" to
ODK
Collect?
Perhaps something similar to the "barcode" type however this would a
launch
a sketch app that would allow the user to draw a picture, add icons,
take a
background picture etc. Presumably the result could be stored and
uploaded
as a media type of image.
If I could chime in, another use case could be sketch maps for household
and/or agricultural surveys. In panel surveys, one needs to find a
household from one wave to the next. In areas where roads aren't named,
like rural Africa, it would be very helpful to sketch the location of the
interviewed household's hut relative to fixed landmarks in the village. The
same logic would apply for a household survey with an agricultural focus.
Not only would one need to locate households from one wave to the next, but
have some graphical means of tracking/referencing their agricultural plots
across rounds (e.g., what did you plant--showing the respondent a map--on
the plot near the road to Thete Trading Center)
As for the needs/constraints/implementation, I was thinking that the sketch
app might be best if, at its core, it were a simple drawing program (i.e.,
as close to paper as possible). Furthermore, this sketch image would only
be useful if it could be called up for panel cases when working through
rounds, so that a map drawn in wave 1 is accessible by another interviewer
in wave 2 when they are trying to locate a household or interview them about
their plots.
Dreaming a bit wildly, this sketch app would be most useful if graphical
elements could be immediately linked to questions. Here I am thinking of a
sketch map of a household's agricultural plots. This sketch could contain
tags/shapes that link to questions--say, a square designates a plot about
which questions could be asked. This set-up would make asking about
agricultural plots very effective--where it would suffice to click on a
plot's icon in the interactive sketch image to ask questions about it.
···
On Mon, Feb 14, 2011 at 8:16 PM, Neil Penman wrote:
Hi Gaetano,
Sounds good!
Perhaps all of those things. I haven't got a detailed set of requirements
for this, however the following uses come to mind:
Take a photo and place a generic marker somewhere on the image. The
question label/hint would provide context for the meaning of the marker.
Sketch a road intersection free style and add markers for POI's.
These markers could be specific to the sketch app and be transparent to ODK
collect.
Add a number of geometric shapes such as squares, lines, perhaps to
show houses and their position relative to a road. Colour code the houses
according to some schema described in the question label or hint.
The sketch capability would presumably be determined by the sketch
application, it may be possible to support multiple installed sketch
applications. Were you thinking of writing your own?
Ideally it should be possible to go back to the question and edit the
sketch. However that may be more complicated to implement then just having
the ability to replace the sketch.
Did you have a timeline for this work? I may be able to find a programmer
to help out if that suits.
regards
Neil
On Tue, Feb 15, 2011 at 11:48 AM, Gaetano Borriello < gaetano@cs.washington.edu> wrote:
Neil:
that is a great idea and along the lines of some things we are talking
about including being able to annotate a photo with arrows, boxes, and
text. Ideally, we will have the ability to have collect call other
phone apps to do specialized input types. If you don't mind, what is
your usage case? Just want to try to understand your
constraints/needs. For examples would you want to be able to edit the
sketch? Use geometric elements or just paint?
Has anybody looked at adding support for a question type of "sketch" to
ODK
Collect?
Perhaps something similar to the "barcode" type however this would a
launch
a sketch app that would allow the user to draw a picture, add icons,
take a
background picture etc. Presumably the result could be stored and
uploaded
as a media type of image.