Inspired by Chris Smith's project this application does classification differently, logs classifications in a local database and maintains an online repository of activities.
The metabolic equivalent of task (MET), is a physiological concept expressing the energy cost of physical activities as multiples of resting metabolic rate (RMR) See http://en.wikipedia.org/wiki/Metabolic_equivalent .
To improve the accuracy of the classifier we have reduced the range of classified activities. A downside of this is that much of the time we can't classify an activity. One of our goals is to measure the level of physical activity undertaken by an athlete. This can be calculated from directly from the triaxial accelerometer data we have sampled. Actical http://www.bmedical.com.au/shop/actical.htm is a commercial product that does this. There are many publications describing the methods and accuracy of using accelerometers to calculate MET. See http://www.ncbi.nlm.nih.gov/pubmed/18002254 , http://www.theactigraph.com/article/research-database/validation-and-calibration-of-physical-activity-monitors-in-children/ . The technique has been tested on a mobile phone and described in the publication, Monitoring Daily Energy Expenditure using a 3-Axis Accelerometer with a
Low-Power Microprocessorhttp://www.unioviedo.es/eminds/index.php?journal=eminds&page=article&op=viewFile&path[]=64&path[]=40 where it is claimed results indicate that this approach can monitor energy expenditure as accurately as a medical calorie meter. In this publication MET is estimated from the classified activity rather than directly from the accelerometer value. When walking speed is used as one of the variables. Pace is calculated and multiplied by stride length but no method for determining stride length is given. In other publications it is stated pace is generally constant and speed is varied by varying stride length during walking.
Add MET as another field to the activity database so that each classification has an additional MET parameter.
On the display include the MET as a numerical parameter following the activity classification description and further classify MET as low medium or high. At the bottom of each bar graph display the numerical MET for the period shown. Also show the MET in the online interface.
The metabolic equivalent of task (MET), is a physiological concept expressing the energy cost of physical activities as multiples of resting metabolic rate (RMR) See http://en.wikipedia.org/wiki/Metabolic_equivalent .
To improve the accuracy of the classifier we have reduced the range of classified activities. A downside of this is that much of the time we can't classify an activity. One of our goals is to measure the level of physical activity undertaken by an athlete. This can be calculated from directly from the triaxial accelerometer data we have sampled. Actical http://www.bmedical.com.au/shop/actical.htm is a commercial product that does this. There are many publications describing the methods and accuracy of using accelerometers to calculate MET. See http://www.ncbi.nlm.nih.gov/pubmed/18002254 , http://www.theactigraph.com/article/research-database/validation-and-calibration-of-physical-activity-monitors-in-children/ . The technique has been tested on a mobile phone and described in the publication, Monitoring Daily Energy Expenditure using a 3-Axis Accelerometer with a Low-Power Microprocessorhttp://www.unioviedo.es/eminds/index.php?journal=eminds&page=article&op=viewFile&path[]=64&path[]=40 where it is claimed results indicate that this approach can monitor energy expenditure as accurately as a medical calorie meter. In this publication MET is estimated from the classified activity rather than directly from the accelerometer value. When walking speed is used as one of the variables. Pace is calculated and multiplied by stride length but no method for determining stride length is given. In other publications it is stated pace is generally constant and speed is varied by varying stride length during walking.
Add MET as another field to the activity database so that each classification has an additional MET parameter.
On the display include the MET as a numerical parameter following the activity classification description and further classify MET as low medium or high. At the bottom of each bar graph display the numerical MET for the period shown. Also show the MET in the online interface.