- Forum
- General C++ Programming
- pattern recognition
pattern recognition
I am working on a program that works as a gesture recognition for sensory glove. Outputs from 14 glove sensors are float numbers ranging from 0(extended) to 1(bent). Every sensor can recognize 4096 positions, but I'm using only two decimals in the float values (0.00 - 1.00) which is totally sufficient for my purpose.
What I've done so far is that I can store sensor values to a file and compare them with actual sensor data and determine if the gesture has been recognized or not. I'm simply comparing each sensor one by one with certain tolerance. The results depend on my ability to recreate the exact position of the sensors with the glove.
I've made some 'secondary' recognition using the sum of all the values from all sensors and comparing both of these values, which could be used with other methods and if all are true, then the gesture would be recognize even if the actual values didn't match the tolerance range of stored values.
Another secondary recognition I've been thinking of is comparing a sequence of numbers created from comparing the sensor values (eg. sensor1 > sensor2 would mean 1, opposite 0; sensor2>sensor3 ....). So the final sequence would look something like this 11011101111110 and these would be compared.
I've been thinking about using neural network for this too, though I'm not sure if I'd be able to do this when I was looking at some neural network codes.
I was wondering if anybody could help me with this either think of some new ways how to recognize the gestures or some sort of iteration algorithm or even neural network.
Thanks in advance
What I've done so far is that I can store sensor values to a file and compare them with actual sensor data and determine if the gesture has been recognized or not. I'm simply comparing each sensor one by one with certain tolerance. The results depend on my ability to recreate the exact position of the sensors with the glove.
I've made some 'secondary' recognition using the sum of all the values from all sensors and comparing both of these values, which could be used with other methods and if all are true, then the gesture would be recognize even if the actual values didn't match the tolerance range of stored values.
Another secondary recognition I've been thinking of is comparing a sequence of numbers created from comparing the sensor values (eg. sensor1 > sensor2 would mean 1, opposite 0; sensor2>sensor3 ....). So the final sequence would look something like this 11011101111110 and these would be compared.
I've been thinking about using neural network for this too, though I'm not sure if I'd be able to do this when I was looking at some neural network codes.
I was wondering if anybody could help me with this either think of some new ways how to recognize the gestures or some sort of iteration algorithm or even neural network.
Thanks in advance
Last edited on
closed account (z05DSL3A)
This may be of interest:
From Wikipedia:
A hidden Markov model (HMM) is a statistical model in which the system being modeled is assumed to be a Markov process with unknown parameters, and the challenge is to determine the hidden parameters from the observable parameters. The extracted model parameters can then be used to perform further analysis, for example for pattern recognition applications. A HMM can be considered as the simplest dynamic Bayesian network.
In a regular Markov model, the state is directly visible to the observer, and therefore the state transition probabilities are the only parameters. In a hidden Markov model, the state is not directly visible, but variables influenced by the state are visible. Each state has a probability distribution over the possible output tokens. Therefore the sequence of tokens generated by an HMM gives some information about the sequence of states.
Hidden Markov models are especially known for their application in temporal pattern recognition such as speech, handwriting, gesture recognition, musical score following, partial discharges and bioinformatics.
http://en.wikipedia.org/wiki/Hidden_Markov_model
From Wikipedia:
A hidden Markov model (HMM) is a statistical model in which the system being modeled is assumed to be a Markov process with unknown parameters, and the challenge is to determine the hidden parameters from the observable parameters. The extracted model parameters can then be used to perform further analysis, for example for pattern recognition applications. A HMM can be considered as the simplest dynamic Bayesian network.
In a regular Markov model, the state is directly visible to the observer, and therefore the state transition probabilities are the only parameters. In a hidden Markov model, the state is not directly visible, but variables influenced by the state are visible. Each state has a probability distribution over the possible output tokens. Therefore the sequence of tokens generated by an HMM gives some information about the sequence of states.
Hidden Markov models are especially known for their application in temporal pattern recognition such as speech, handwriting, gesture recognition, musical score following, partial discharges and bioinformatics.
http://en.wikipedia.org/wiki/Hidden_Markov_model
Last edited on
Topic archived. No new replies allowed.