Challenge file: https://github.com/mlesterdampios/tcon8/tree/main/intersection
In this OSINT challenge, we’re given just a single photo of an intersection—and unfortunately, it contains no useful metadata to lean on.
My approach was a little unconventional: I cropped the image into smaller, “searchable” chunks and treated each piece like its own clue.
Attempt 1: The street sign
I started with the most obvious lead—the sign that clearly shows “4th Street”, plus a few characters that were too blurry to read. I ran it through reverse image search, but the results were way too broad and didn’t give a clean match.
Attempt 2: Brute-forcing Google Maps
Next, I tried searching Google Maps for places with “4th Street.” That idea died fast—there are tons of 4th Streets, and manually checking them felt like trying to find a specific grain of sand on a beach.
Attempt 3: Narrowing by environment
To reduce the search space, I tried filtering my mental shortlist using environmental hints—especially the pine trees, which suggested a region where they commonly grow. It helped a bit, but the results were still massive.
At that point, I also tried “walking” around in Google Maps/Street View and doing a few more reverse image searches… but everything still felt like loose threads and dead ends.
The breakthrough
Here’s what finally worked:
Instead of focusing on the street sign, I reverse-image searched the building across the intersection.
And boom—this produced promising, specific results.
From there, it was mostly trial and error: the challenge required the building name in an exact format, so once I had candidate matches, I just tested variations until I hit the correct submission format.
