Computer vision-assisted analysis of motion of a person

1 . A method performed by a motion tracking system, the method comprising:

capturing, by an optical device of the motion tracking system, a calibration image of a person while the person is in a predetermined calibration position;

processing, by at least one computing device of the motion tracking system, the calibration image to determine at least two calibration length values based on distances between anatomical landmarks associated with at least one body member in the calibration image;

capturing, by the optical device, an exercising image of the person while the person is in an exercising position corresponding to a predetermined exercise;

processing, by the at least one computing device, the exercising image to determine at least two target length values based on distances between the anatomical landmarks associated with the at least one body member in the exercising image;

determining, by the at least one computing device and based on the at least two calibration length values and the at least two target length values, at least one depth-related parameter associated with a target body member of the at least one body member, wherein the at least one depth-related parameter comprises a range of motion in depth of a first anatomical landmark of the target body member relative to a second anatomical landmark of the target body member, and wherein the depth is in a direction perpendicular to two other directions corresponding to two dimensions captured by the optical device in the calibration image and the exercising image;

using, by the at least one computing device, the at least one depth-related parameter to track motion of the person with respect to the predetermined exercise; and

determining, by the at least one computing device, whether the person has moved the first anatomical landmark relative to the second anatomical landmark according to at least one predetermined range of motion in depth associated with the predetermined exercise.

2 . The method of claim 1 , wherein the first anatomical landmark is in a position relative to the second anatomical landmark that differs between the calibration image and the exercising image, and the at least one depth-related parameter further comprises a difference in the depth of the first anatomical landmark relative to the second anatomical landmark.

3 . The method of claim 1 , wherein the using of the at least one depth-related parameter to track the motion of the person comprises:

comparing, by the at least one computing device, the range of motion in depth with the at least one predetermined range of motion in depth associated with the predetermined exercise.

4 . The method of claim 1 , wherein the at least one body member comprises a first body member and a second body member, wherein the at least two calibration length values comprise, for each body member of the first body member and second body member of the person, a calibration length value based on a distance between the anatomical landmarks that are associated with the body member in the calibration image, wherein the at least two target length values comprise, for each body member of the first body member and second body member of the person, a respective target length value that is based on a respective distance between the anatomical landmarks that are associated with the body member in the exercising image, and wherein each respective target length value corresponds to the distance between the same anatomical landmarks used for determining the distance of each corresponding calibration length value.

5 . The method of claim 4 , wherein the target body member and the second body member are the same body member, and in the predetermined exercise the first body member is not to be moved and the second body member is to be moved.

6 . The method of claim 5 , wherein determining the at least one depth-related parameter associated with the target body member comprises:

determining, by the at least one computing device, a length variation factor based on the calibration length value of the first body member and the target length value of the first body member;

determining, by the at least one computing device, a scaled length value associated with the second body member based on the calibration length value of the second body member and the length variation factor; and

determining, by the at least one computing device, the at least one depth-related parameter based on the scaled length value and the target length value of the second body member.

7 . The method of claim 6 , wherein determining the at least one depth-related parameter comprises calculating a trigonometrical relationship between the scaled length value and the target length value of the second body member.

8 . The method of claim 4 , wherein in the predetermined exercise the first body member stays substantially motionless, and the second body member is to be moved such that at least a first anatomical landmark of the second body member moves relative to a second anatomical landmark or such that the second body member rotates relative to the second anatomical landmark.

9 . The method of claim 1 , further comprising:

capturing one or more additional exercising images; and

determining the at least one depth-related parameter associated with the target body member for each of the one or more additional exercising images.

10 . The method of claim 1 , wherein the at least two calibration length values are calibration length values of a first body member of the person, and the at least two target length values are target length values of the first body member of the person, each target length value corresponding to the distance between the same anatomical landmarks used for determining the distance of each corresponding calibration length value, and wherein the at least one depth-related parameter is determined based on a first target ratio and a second target ratio, the first target ratio being between the at least two calibration length values of the first body member and the second target ratio being between the at least two target length values of the first body member.

11 . The method of claim 1 , wherein the at least one depth-related parameter indicates the direction as either towards the optical device or away from the optical device.

12 . The method of claim 11 , wherein the direction is based on at least one biomechanical constraint of the target body member.

13 . The method of claim 11 , wherein the direction is based on the predetermined exercise.

14 . The method of claim 1 , wherein the optical device is at a same location and has a same orientation when capturing the calibration image and when capturing the exercising image.

15 . The method of claim 1 , wherein the motion tracking system comprises a motion tracker arranged on the person, the motion tracker including an inertial measurement unit, and the method further comprising:

processing measurements of the motion tracker taken while the person is in the predetermined calibration position and while the person is in the exercising position, thereby obtaining a set of measurements at different time instants;

detecting that the at least one depth-related parameter and the set of measurements fulfill one or more predetermined constraints associated with the predetermined exercise; and

determining, based on the detection, that the predetermined exercise has been correctly executed.

16 . The method of claim 1 , wherein the motion tracking system comprises a motion tracker arranged on the person, the motion tracker including an inertial measurement unit, and the method further comprising:

processing measurements of the motion tracker taken while the person is in the predetermined calibration position and while the person is in the exercising position, thereby obtaining a set of measurements at different time instants;

detecting that the at least one depth-related parameter and the set of measurements do not fulfill one or more predetermined constraints associated with the predetermined exercise; and

determining, based on the detection, that the predetermined exercise has not been correctly executed.

17 . The method of claim 1 , further comprising:

providing, by the motion tracking system and prior to the capturing of the calibration image, at least one user-perceptible signal indicative of the predetermined calibration position to instruct the person to assume the predetermined calibration position.

18 . The method of claim 1 , providing, by the motion tracking system and prior to the capturing of the exercising image, at least one user-perceptible signal indicative of the predetermined exercise to instruct the person to perform the predetermined exercise.

19 . A motion tracking system comprising at least one computing device and an optical device, the motion tracking system being configured to perform operations comprising:

capturing, by the optical device, a calibration image of a person while the person is in a predetermined calibration position;

processing, by the at least one computing device, the calibration image to determine at least two calibration length values based on distances between anatomical landmarks associated with at least one body member in the calibration image;

capturing, by the optical device, an exercising image of the person while the person is in an exercising position corresponding to a predetermined exercise;

processing, by the at least one computing device, the exercising image to determine at least two target length values based on distances between the anatomical landmarks associated with the at least one body member in the exercising image;

determining, by the at least one computing device and based on the at least two calibration length values and the at least two target length values, at least one depth-related parameter associated with a target body member of the at least one body member, wherein the at least one depth-related parameter comprises a range of motion in depth of a first anatomical landmark of the target body member relative to a second anatomical landmark of the target body member, and wherein the depth is in a direction perpendicular to two other directions corresponding to two dimensions captured by the optical device in the calibration image and the exercising image;

using, by the at least one computing device, the at least one depth-related parameter to track motion of the person with respect to the predetermined exercise; and

determining, by the at least one computing device, whether the person has moved the first anatomical landmark relative to the second anatomical landmark according to at least one predetermined range of motion in depth associated with the predetermined exercise.

20 . The motion tracking system of claim 19 , further comprising a motion tracker being adapted for arrangement on the person, the motion tracker including an inertial measurement unit.