Detection Of Finger Movement Based On Three-axis Accelerometer

The detection of human hand movement parameters can be applied in many aspects. Such as the assessment of hand motor function, physical rehabilitation therapy coupled with electrical stimulation and so on. There are many different ways to detect the hand movement, from the earliest goniometer, magnetic sensor, ordinary camera technology to subsequent ultrasound technology, high-speed camera technology and so on. However, there are some disadvantages in these methods such as easily influenced by external interference, impeding the movement of fingers, huge amount of data and the high cost of equipments. This paper select three-axis accelerometer named ADXL330 based on MEMS to detect the parameters of finger movement. All the disadvantages mentioned above can be avoided when this sensor is used to detecting the finger movement. And there are some advantages in this method such as more information can be attain from static state to kinematic state, from plane location to space location.On the basis of understanding of the peripheral circuits of the sensor, the detecting devices including circuit board of sensor and the fixture were made. LabVIEW and its hardware were used to acquisiting data and Matlab was used to process the data. The data-processing procedures include data filtering, smoothing, computing and graphics rendering. On the basis of analyzing the inner configuration of the sensor and experiments, the sensor measurement principle mathematic model was confirmed and angles between the direction of gravity and three axis of the sensor were selected as detection parameter. Different experiments were done to verify the correctness of this mathematic model. In the demarcate experiment fixing a sensor on a steel bars which was put on different angles in the vertical plane, the results show that the absolute errors of sensor's static state output are less than 4%; in the joint model experiment fixing two sensor on the model which was put on different angles in the vertical plane, the results show that the errors of two sensor are in a certain range. These two experiments show that one sensor can be used to detect the angle between knuckle and the direction of gravity, two sensors can be used two detect angle between two different knuckles. In the forearm experiment fixing a sensor on the vicinity of the wrist when it moves, at the same time, high-speed camera was used to record the process of forearm movement, the results of sensor was contrasted with the results of high-speed camera, which shows that the traces of them are same and the absolute errors of the final value of angles are less than 5%, but certain time lag is exist.In order to verify the accuracy of the model further, step motors and the rigid material were used to make the model of finger. And the results of the experiments fixing sensors on this motor-model which was put in different angles in the vertical plane, the results show that the sensors have good repeatability and accuracy. Therefore, all the results of experiments verify that the sensors can be used to detect the movement of fingers.In order to use the sensors to detecting finger movement, four sensors were used to detect different volunteers'finger movements (index finger). Experimental data acquisition system is comprised of LabVIEW software and acquisition hardware named NI9205 module. Before the experiment, the motion modes were prescribed and the volunteers were trained. In the experiments, volunteers were requested to move their index fingers keeping two joints of their fingers static first, and then fisting naturally. The outputs of the sensors were processed which contain the information of angle, location, angle velocity and acceleration. Preliminary experimental results show that the moving sequence of the knuckles of the index finger is different, the moving speed and acceleration of them are different.All the experiments show that the angles between axis of sensors and the direction of the gravity can be attained when three-axis accelerometers are used to detect the parameters of movement, from which more movement information can be attain. The most important thing is that this method provides the basis of theory and experiment for space location of objects.