VisionBased Motion Sensor Jacaqueta Davis Michael Ledet Department

Vision-Based Motion Sensor Jacaqueta Davis Michael Ledet Department of Computer Science and Industrial Technology Southeastern Louisiana University 500 W University Ave, Hammond, LA 70402

Objective To build a low cost motion detection device that will detect a human presence.

Initial Constraints 1. Low cost components 2. Advanced Capabilities for Interfacing

Initial Constraints → Results 1. Low cost components → Components combined price under $20 2. Advanced Capabilities for Interfacing → Can program in Arduino IDE and AVRdude (command line terminal)

Main Components Adafruit Trinket Mini • Adafruit Trinket Mini • Pinhole camera • Additional standard electronic components Pinhole camera

Outline • Design test circuit • Practical interfacing with microcontroller and camera

Outline • Programming for: - Successive snapshots - Data Capsulation - Image analysis • Troubleshooting/debugging - (Ongoing)

Data Capture ● Due to processing power of hardware: - Utilize minimal amount of data required to achieve motion detection

Data Capsulation ● Due to processing power of hardware: - Utilize minimal amount of data required to achieve motion detection

Image Analysis Comparison of two subsequent snapshots from the video feed: vs. Data Output

Progression • Initial interfacing with camera • Scripts: - Snapshot capture - File handling/Data Capsulation

Obstacles • Processing information locally -Remote computer for data analysis? • Tuning for sensitivity -Detection for people vs anything moving in the frame

Deliverables

Deliverables

References Trinket Mini 5 V: https: //www. adafruit. com/products/1501? gclid=CK 68 g. PKMh 9 MCFdy 6 w. Aodk 3 MF 6 Q Optical Sensor: https: //www. adafruit. com/products/3202 https: //en. wikipedia. org/wiki/PIC_microcontroller#/media/File: Microchip_PIC 24 HJ 32 GP 202. jpg https: //i. ytimg. com/vi/m-_t_a. Il. Qwg/maxresdefault. jpg