Image-to-text technologies, currently used for their efficiency at creating electronic copies of books, documents, exams, and texts. In today's mobilized world, new applications for old technologies open up, allowing people to approach old problems in new ways. "DroneText" is an applications that seeks to integrate image-to-text with drone technologies for finding their lost or stolen cars.
With DroneText, the user inputs the license plate number of his or her car, and can then activate the drone to move it in a variety of directions - forward, backward, left, right, up, down, and anywhere in between. The drone can take photos, and then process these photos to decipher any text found - both alphabetic and numeric. Once the text has been deciphered, its similarity to the license plate number can be ranked on a scale between 0.0 and 1.0. If the similarity is great enough, the drone's location is sent to the user, allowing him or her to track down his stolen car. This will make it significantly easier to locate lost cars that lack GPS systems.
Though the current intent is to help track lost cars, DroneText can be used to discover the location of any text.
DroneText was created using C++, Autoflight, and Microsoft's OCR API.
Log in or sign up for Devpost to join the conversation.