In this article, we propose Fido, a fuzzy-based string comparison method for extracting content from data lakes in a camera-based Internet of Drones (IoD) environment. Data lakes support data dumping in their native format, which makes them suitable for real-time deployments such as IoD. However, parsing through these unstructured databases has its concerns, particularly data extraction. Existing works on image-based content extraction focus on complex geometric and matrix mathematical operations, which are expensive in terms of both computation and time. A straightforward and fast solution is necessary for overcoming such challenges. Toward this, we limit our operations to string comparisons between the user's request and tags in the data lake. In particular, we adopt a fuzzy-based approach as it is capable of efficiently handling spelling variations, out-of-order words, and partial matchings, compared to conventional string comparison methods. Through lab-scale experiments, we demonstrate the efficacy of Fido with an almost 80 percent similarity ratio between the request and response images and processing time of 80 µs. Further, we observe minuscule deviations between the coordinates of the request and response images (0.005 units).