How to Improve IR LED Population Counter Accuracy for Groups or Children Detection?

The discussion addresses improving the accuracy of an IR LED-based population counter designed to detect people entering or exiting by monitoring the sequence of beam interruptions. A key challenge is accurately counting groups or individuals carrying infants, as the current setup registers them as a single person. Suggestions include acknowledging the limitations of simple IR beam systems for complex scenarios like side-by-side walkers or people carrying children, and documenting these constraints in the project report. More advanced solutions involve integrating CCTV cameras with image processing software, though this adds significant complexity. Alternative hardware enhancements proposed include adding diagonal IR beams at angles (e.g., 45 degrees) alongside orthogonal beams to better differentiate between single individuals and groups or objects like strollers. Measuring IR signal strength to detect anomalies or suspect readings and incorporating mechanical elements like turnstiles to aid separation were also discussed. Statistical adjustments (fudge factors) based on the frequency of special cases were suggested to improve count accuracy. Overall, the consensus is to balance project complexity with achievable goals, possibly using multi-angle IR beam arrays and acknowledging inherent limitations in the final design.
AI summary based on the discussion. May contain errors.