Home > About
Home > About
Below is a brief overview of what WISDAM does. The software is based on our experience of developing imagery / drone survey methods and our intrinsic understanding of the data requirements for wildlife surveys.
Advancements in the capabilities of imaging technology and drones (UAVs / UASs) have enhanced our capacity to survey wildlife abundance, distribution and habitat use, and improved the accuracy of the data collected. Yet these surveys result in large image datasets that can take hundreds of hours to review manually. We developed our WISDAM software (Wildlife Image Survey – Detection and Mapping) to support drone or piloted-aircraft imagery surveys by standardising and automating the image review process.
When you import your images and select the appropriate ‘data loader’ according to the aerial platform you’ve used (e.g. select your drone model), your images are immediately georeferenced to real-world coordinates. Image metadata can then be exported as csv or GIS file types.
When you import your images, you can add attributes such as flight references, transect numbers, or survey block.
You can design your own project metadata attributes according to your unique survey needs. Create your own environmental categories to classify your images, and set up your own variables to describe each animal sighting.
You can ensure each image is reviewed using standardised methods – set your zoom and use WISDAM’s grid navigation to walk through your image section by section.
You can easily detect and assign resights of the same individual animal using WISDAM’s projection feature, which displays the location of the bounding box of animals already labelled as they should appear in overlapping images. Assign multiple sightings of a single animal to a resight set to ensure the animal is only counted once.
You can score each of your images according to your custom environmental categories, and these scores are propagated to subsequent images until changed. The scores are then automatically assigned to each sighting, but can be adjusted for each animal sighting in an image if needed.
You can view all of your sighting using WISDAM’s gallery page, which allows you to filter and sort the sightings according type, detection method, resight set and more. Click on each sighting to review and edit it’s attributes, or to go to the full image.
You can review your images and sightings using WISDAM’s GIS page and tag multiple images with additional metadata such as survey transect. You can also automatically group sightings according to proximity to assist with detecting resights of individual animals.
You can visualise, verify and manage the output from an automated (AI) detection model by importing them via WISDAM’s AI page. Scroll through the detections and highlight the true positives before importing these detections into your project dataset. The detections can be filtered according to the probability assigned by your AI model, as well as classifications (e.g. animal type) or AI run number.
You can compare AI sightings with manual sightings, or compare multiple manual reviews using WISDAM’s compare page. Sightings from multiple sources are automatically matched according to pixel coordinates and the matches are displayed for you to confirm. Save these comparisons to review at a later date, or export the comparisons as a csv file. These comparisons are designed to be used to calculate perception bias. The exported data consists of one row per individual animal with comparative data for each detection source.
Your image and sighting data can be exported as csv files or GIS files (shapefiles, json, kml). You can also export an image AI training dataset – all images containing sightings are automatically compiled into one folder, along with folders of sighting thumbnails for each type (e.g. each taxa) and a label text file.