The process of analyzing worldwide data on animals is getting easier, and it all comes down to data harvesting techniques that are becoming more accessible the scientists thanks to AI systems.
With the urgency of many species threatened around the world, having real-time data on the nature of endangered species can be crucial to ensuring their survival. As biologists are not always able to work on-site or collect specimen data, it’s often difficult to access real-time information on some of the Earth’s most threatened species.
One of the most traditional ways of tracking numbers in biology is doing satellite photos or aerial counts. The problem with many of these wide-angle photos is that it takes time and resources to spot the animals in this footage. It often falls to the researcher to go through the data and physically count the animals in the aerial survey. Rather than utilizing these older methods is becoming easier for biologists to get their accounts than ever before.
Using some of this data and imaging software, it’s possible for biologists to use a computer model to track numbers on their aerial counts today. While Wildbook, a software program developed by Portland-based conservation tech nonprofit Wild Me, for example, can automatically identify individual animals based off of their coat patterns, your outlines, and other features, even from the air. Having access to tools like this can help to make sure that biologists can continue to go through data at a much faster rate. In one of the first tests of this software the San Diego zoo did a study on African reticulated giraffes in northern Kenya. While Wildbook was able to instantly capture data from photographs that were taken over two days, delivering accurate results on the current population of giraffes in the area of the national park that was surveyed. These types of results would typically take weeks or even months of study to produce before.
What’s planned is to eventually make the software available to park rangers and tourists in the park. This can mean that any of the photos uploaded from the area could be of use to biologists creating an even larger data set to reflect the population.
Having access to more of these large data sets in threatened populations could have worldwide tourists, biologists, park rangers and amateur conservation list all working together to gather quality data that can be used in conservation efforts.
AI systems are also leading to the analysis of this data becoming a much faster process. Going through millions of trap photos for example with something that would commonly take image enhancement specialists, biologists and researchers. Today, there are AI systems that can work on every facet of this information to go through millions of photos over just a short amount of time and at a fraction of the cost. The banks of data that have been collected and left unanalyzed can finally be properly put to use.
The software can only continue to get better, and there are even elements in the works that could one day identify individual animals in threatened populations. Current tools are currently in place with whale watching technology that can “fingerprint” whales by examining their tails in image data.
These types of improvements are offering us a brand-new way that we can look at animals and conservation worldwide. As this technology improves, we could see improvements in conservation efforts worldwide as a result of the efforts here.
Source: The information used in this article is from https://www.nationalgeographic.com/animals/2018/11/artificial-intelligence-counts-wild-animals/