Senno Gait analyzes humans’ gait using the micro motion sensors. By capturing and detecting the foot motion and impact, based on the correlation between foot motion and pressure distribution, pressure distribution like center of pressure can been computed based on motion sensor data. Our model has been verified through long-term researches. The specific theory is illustrated below.
The points on the bottom surface of a foot are not completely independent. Instead, they related to each other to a certain degree. Based on the face that all the points must be set on the human-sole-shaped surface, they follow the theory of MANIFOLD. Consequently, we can derive global structure through local structure.
The foot pressure range and distribution are related to the motion of the foot. Physicians can provide many examples, such as, when the toe rises, the foot pressure moves towards the heel. When inversion happens, the foot pressure shifts to the out-side part. Also, when the foot has the tendency of eversion, it tends to be accompanied by the arch collapse and the and more foot pressure of out front. When the impact of the bottom is big, the pressure on the bottom is big as well, etc. But all this knowledge is from practical experience and still not systematic. A better method is to abstract quantified models using big data techniques.
Sennotech has accumulated more than 100,000 gait data entries since 2009, which can be used to do the cloud computing to analyze the big data. Details are described below.
One method of data collection was using our wearable device with highly integrated pressure sensor and motion sensors, shown in the following picture. In the hundreds of thousands of data we’ve collected, we have 10% from America and 90% from China, and the data are distributed through various health statuses, diseases and ages. These data are a key part of big data AI tech evolution.
The aforementioned data-collection device integrated motion sensor, foot pressure sensor, can get motion data and pressure data at the same time (The method can adjust the data of different sensors and sample interval to the same time space so as to extract feature and get the fusion result. ) .
And because of its portability, it can facilitate the acquisition of daily life scene data. A sense of science and technology has accumulated a complete gait database by this method.
The technology is based on our gait database, utilizing the technology of “Mainifold Learning”, “SVM”, “high precision motion tracking”. The technical process is as follows:
Generally, technology is divided into two parts: training phase and calculation phase. In learning phase, we performed a structurized learning process over the database, to make sure that relationship between motion and pressure of any gait type or in any scenario can ben learnt. Thus we are able to build a ‘Motion-Preure association model’. This model can be used in our cloud online computing, to rebuild and reconstruct the pressure data from the user equipped with our motion sensor.
Use Senno Gait to get pressure distribution