The Possibilities of Image Recognition

After the tragic Maryland shooting transpired at the Capital Gazette, in late June, police were desperate to find the culprit. Without cooperation from the suspect and an inability to match fingerprints, the police department turned to a system including millions of photos of criminals as well as people who carried driver’s licenses with the state: the Maryland Image Repository System (Brandom). Thanks to this database (MIRS), police were able to identify Jarrod Ramos as the shooter. Law enforcement could lean on image recognition, when all customary methods failed, and Maryland was able to see “a huge win” for justice (Brandom).

Image recognition is by no means perfect. Especially when it comes to governmental and law enforcement uses, there are still many civil liberties to consider and restrictions to place over the course of time. However, in terms of image recognition, the potential for good is undeniable and the possibilities are ample. Here are some basic statistics for you. In less than 3 years, 10 million self-driving cars are predicted to be in use (How Image Recognition..). Companies like Tesla, Ford, Uber, and Google are currently perfecting their image recognition systems to maximize safety, productivity and efficiency in these high-tech vehicles. Additionally, Google has used over 120,000 medical images to train its machines to utilize image recognition. Their technology was able to identify mis-diagnoses in diabetic retinopathy 10% more accurately than human doctors. And by 2020, image recognition is anticipated to become nearly a 30 billion dollar industry (How Image Recognition..).

Still, that only covers a few of many current uses of image recognition. The technology is also making progress in the gaming industry, the medical field, in tourism, and education (Golemanova). In fact, it is even getting as niche as affecting food packaging. The Multi-Color Corporation and Talkin Things are working on interactive image recognition packaging with a new ale beer. According to reports, this ale will contain a label that can be scanned onto a mobile application. An image on the label will then appear in the app and precede to use facial recognition to detect a consumer’s mood, and use this basis to interact with the user (Haigh). By utilizing AR, and of course, image recognition, this packaging technology could mean a complete transformation in the food industry, giving customers a more personalized and entertaining experience.

Whether it’s large or small, image recognition has involved itself in many different walks of life, and it will only grow from here on out. Be prepared for many more talking beer labels and self-driving cars in your future!