Courtney Nadine Coleman
The Baby Track App
Tracking gross motor development in infants by using the power of cameras.
http://babytrackapp.tumblr.com
Classes
Biomechanics for Interactive Design,The Nature of Code Studio
Biomechanics for Interactive Design,The Nature of Code Studio
This was a research-based exploration of how observing the rotational and angular motion of joints in infants at an early age over time can generate data that can be used to determine whether or not there is neuromuscular dysfunction in an infant. The data can be as a means for a doctor to possibly determine if there are developmental delays or developmental disability such as cerebral palsy, down syndrome, autism, and many others.
Audience
Parents, doctors, caretakers, and other supporting professionals that work with infants and toddlers.
User Scenario
A parent can use a web cam on their computer to shoot their infant moving. From there data regarding the amount of movement in the form of pixel changes are generated and displayed to the parent, numerically and visually.
A parent can attach LEDs to a child's hip, knee, and foot, turn on the web cam, run the processing sketch, and automatically record, while seeing the angle of flexion, extension over time.
Implementation
It was made using processing and a web cam. The markers are colored LEDs and these are turned into vectors. From the use of these vectors, we are able to get the angles.
Conclusion
From a research perspective, I learned a significant amount about gross motor development and about the biomechanical implications of how analyzing motion can help determine if there is a neuromotor dysfunction.
From a technical standpoint, I learned about the power and limitations of a variety of computer vision tools and how they can be used to get biomechanical data such as measuring the angles of flexion and extension. I initially began with the Kinect and attempted to track a baby's skeleton that way. However, I learned that the Kinect will only track a person's body that is sitting or standing upright and is well above 40 inches tall. Next, I used AR markers and I learned that AR markers work best when bodies and surfaces are still. Since I wanted to be able to track the motion of an infant this was not idea. After that, I moved on to the motion detection image processing technique. While I was able to obtain angular motion data, since there is no way to distinguish what is a leg, versus a knee, and a hip. However, the motion data obtained from this could still be useful, but I wanted to get more specific. Finally, I opted to use the color tracking and a wearable type of system to basically place colors on the child that correspond with the child's hips, knees, and foot placement.With color tracking, there are still some limitations. For example, the child should be in a room with ambient, no harsh or extreme lighting. They must be placed on a flat, solid colored surface with a very little background distractions in order for the camera not to be confused. I placed "shades" over the LEDs to help the light spread less so it can be as specific as possible and ultimately the LEDs will be placed on wearable bands so they can easily be placed on the grown/hip, knee, and foot areas of the body.
Audience
Parents, doctors, caretakers, and other supporting professionals that work with infants and toddlers.
User Scenario
A parent can use a web cam on their computer to shoot their infant moving. From there data regarding the amount of movement in the form of pixel changes are generated and displayed to the parent, numerically and visually.
A parent can attach LEDs to a child's hip, knee, and foot, turn on the web cam, run the processing sketch, and automatically record, while seeing the angle of flexion, extension over time.
Implementation
It was made using processing and a web cam. The markers are colored LEDs and these are turned into vectors. From the use of these vectors, we are able to get the angles.
Conclusion
From a research perspective, I learned a significant amount about gross motor development and about the biomechanical implications of how analyzing motion can help determine if there is a neuromotor dysfunction.
From a technical standpoint, I learned about the power and limitations of a variety of computer vision tools and how they can be used to get biomechanical data such as measuring the angles of flexion and extension. I initially began with the Kinect and attempted to track a baby's skeleton that way. However, I learned that the Kinect will only track a person's body that is sitting or standing upright and is well above 40 inches tall. Next, I used AR markers and I learned that AR markers work best when bodies and surfaces are still. Since I wanted to be able to track the motion of an infant this was not idea. After that, I moved on to the motion detection image processing technique. While I was able to obtain angular motion data, since there is no way to distinguish what is a leg, versus a knee, and a hip. However, the motion data obtained from this could still be useful, but I wanted to get more specific. Finally, I opted to use the color tracking and a wearable type of system to basically place colors on the child that correspond with the child's hips, knees, and foot placement.With color tracking, there are still some limitations. For example, the child should be in a room with ambient, no harsh or extreme lighting. They must be placed on a flat, solid colored surface with a very little background distractions in order for the camera not to be confused. I placed "shades" over the LEDs to help the light spread less so it can be as specific as possible and ultimately the LEDs will be placed on wearable bands so they can easily be placed on the grown/hip, knee, and foot areas of the body.