ML
AutoClicker
A machine-learning desktop tool for repetitive clicking work. It reads numbers from a chosen area of the screen with OCR and ML-based digit detection, then clicks precisely in a target area.
- Role
- ML model development and OCR integration
- Stack
- Python · OCR · Machine learning
01 ProblemRepetitive, number-driven clicking tasks are slow and error-prone by hand.
02 ApproachRead values from a screen region in real time, then trigger precise clicks.
03 TechnologyPython, OCR and ML-based number detection in a desktop app.
04 OutcomeConfigurable automation, later extended with image recognition and voice.
Some desktop workflows come down to watching a number on screen and clicking in response, over and over. Done by hand, it is slow, tiring and easy to get wrong.
01Read the screen in real timeOCR combined with an ML number-detection model extracts values from an input region the user marks on screen.
02User-defined regionsUsers choose their own input and output regions, so the same tool can automate many different tasks without code changes.
03Extend beyond numbersAn extended version adds image recognition and voice processing, so automation can react to visual and spoken cues as well.
- Language
- Python
- Vision
- OCR with ML-based number detection
- App
- Desktop application with custom input and output regions
- Extended
- Image recognition and voice processing
