Best way to present large number of tiff files in portal - or images in general

sophered · ‎11-05-2024

Hello! I need some guidance on working with images in portal. I have over 500 image files that I have to georeference (which is the easy part) and then display on a map. The issue is that these images have associated information that MUST be included to allow for filtering/querying in some sort of web application. I have not worked much with raster formats so I am unsure where to start with this. Some questions that I have are:

Can I somehow add attributes to any of the raster formats that are accessible?

Is tiff the best format to have the images in currently?

Data stores? Image Server? I dont even know where to start with that one!

I have experience with web applications but NO experience with raster formats so my main issue is with that. I am pretty comfortable with working with python so that is also an option if there is some type of workaround. Any suggestions are welcome and greatly appreciated!

TonyContreras_Frisco_TX · ‎11-05-2024

I would recommend using a mosaic dataset to store the georeferenced image files. These can be stored in either a file GDB or Enterprise GDB and will require an Image Server license to include in a map or image service. As far as using TIFF images, they are a good lossless format and commonly used for remotely sensed imagery to be displayed in maps. Because they are lossless and usually very high quality images, the file sizes can be large and the time to render them long. The benefit of using a mosaic datasets is that you can generate overview images, which function similar to the old raster pyramids from back in the day. Those are generated once and stored in a long-term location. You can alter the raster format and compression used for the overviews to suit your purpose in the Properties.

I would try exporting the footprint feature class from the mosaic dataset (contains polygon outline of each TIFF added with attributes used for MD functionality) and updating the schema to match the attributes you want to manage. When exporting, only query features with the 'Primary' category. The new feature class can be laid over the imagery (with blank symbology) and allow pop-ups or whatever functionality you need to display or convey the other attributes you are including related to each image. Wait to do this after you have the Mosaic dataset the way you want it.

sophered · ‎11-06-2024

@TonyContreras_Frisco_TX thanks for the quick response! would this method be able to filter/ query? im supposed to attach a date field to the individual images and be able to display different images based on year. i also have MANY images, some overlapping, but they are from different years? would a mosaic dataset allow me to have many individual images that can be queried by this year field? would it be in the form of MANY different mosaic datasets or just one?

Ed_ · ‎11-06-2024

For time series raster mosaic data you should look into

Create a multidimensional mosaic dataset from a set of time series images

Question | Analyze | Visualize

sophered · ‎11-18-2024

@TonyContreras_Frisco_TX @Ed_
Thank you for your quick responses. After looking into the mosaic dataset, it seems like I would need the image server extension, which we unfortunately do not have. Are there any alternatives?

Ed_ · ‎11-18-2024

A free alternative could be using the Terra package in R programming language. You should be able to bring that raster stack into Pro.

I am sharing a link on how to create a timeseries raster in R that might be helpful