Ideas
- Tv/Car embedded surveillance.
- Hidden camera on a tv/car.
- Detects when powered on outside of a defined geofence and used by an unknown person.
- Capture on demand/automatically images on predefined actions.
- Uploads the captured pictures to the cloud/server.
- Outdoor/Indoor surveillance.
- Detects familiar/unknown persons.
- Capture on demand/automatically images on predefined actions.
- Uploads the captured pictures to the cloud/server.
- Send notifications to the configured recipients.
- Garage door automation.
- Detects automatically the license plate and opens the garage door.
- Uploads logs with the activity.
- Present the activity logs over time.
- Send notifications to the configured recipients.
- Water level surveillance/monitoring.
- Send notifications if detects water over a certain level.
- Capture the current levels.
- Present historical data over time.
- Send notifications to the configured recipients.
- Air quality monitoring.
- Notify automatically when the configured threshold values are hit.
- Capture the current levels.
- Present historical data over time.
- Send notifications with the current values to the configured recipients.
- Apidictor.
- An apidictor is an instrument which measures and records the sound in a beehive. The instrument records the aggregate sound made by the buzzing of the bees' wings. They were thought to be useful for predicting when a colony is preparing to swarm. E.F. Wood invented and patented the apidictor in 1964.
- Using TensorFlow, train a model with two microphones, one attached to the hive to capture the hive activity and the other to capture the outside noise. After recording for a few seconds, eliminate the outside noise, the app should be able to draw a conclusion if the hive is preparing to swarm or not.
- https://web.archive.org/web/20050717233655/http://www.beesource.com/plans/apidictor.htm
- Noise level monitoring.
- Detects automatically if the max threshold level was reached.
- Sends alerts to the configured user.
- Define separate threshold values for different time intervals.
- Road condition detection.
- Detects automatically the road conditions, like normal, wet, muddy, covered by leaves, etc.
- http://liu.diva-portal.org/smash/get/diva2:532767/FULLTEXT01.pdf
- Road sign detection.
- Detects road signs like stop, speed limit, etc.
- https://ieeexplore.ieee.org/abstract/document/4220659
- https://ieeexplore.ieee.org/abstract/document/1232697
- Mood detection.
- Train using TensorFlow a model capable of recognizing the moods of a person.
- Define actions based on the detected mood.
- https://ieeexplore.ieee.org/abstract/document/4543754
- https://www.tandfonline.com/doi/abs/10.1080/02699939508408982
- Estimate the number of persons in a room.
- Using the video feed or static images detect the number of persons in a room.
- Train the model using TensorFlow.
- Record the values over time.
- Present historical data over time.
- Weather Station.
- Collect temperature/humidy/etc values in a cloud DB.
- Analyze the collected data.
- Present realtime and historical data over time.
- https://www.raspberrypi.org/blog/build-your-own-weather-station/