Real-Time Smart Categorization System with OCR and AI
Budget: $30 – $250 USD
I want to develop a Python-based system that, while mirroring the mobile app screen opened via Bluestacks or Samsung Flow to my computer using scrcpy-win64-v3.3.1, processes the live stream with modules such as PaddleOCR, OpenCV, and Gemini 2.0 Flash. This system will use the headers I have defined in the header library to process the data in real time, categorize it intelligently, and make it continuously trackable. I also have partially written but non-working code created with Cursor Web, which can be used as a reference. The goal is to keep the system simple but functional.
From the user’s perspective: I open the page and scroll from top to bottom and then back from bottom to top. The data on the page is matched with the headers in the header library and categorized accordingly. I want a system that works with 100% accuracy.
The project has a very simple system. I believe it can be completed within a few hours.
From the user’s perspective: I open the page and scroll from top to bottom and then back from bottom to top. The data on the page is matched with the headers in the header library and categorized accordingly. I want a system that works with 100% accuracy.
The project has a very simple system. I believe it can be completed within a few hours.