SkillSpector introduction

Introduction to Motion and skill Analysis with SkillSpector

Video based movement tracking/analysis is composed of five simple steps.

Step 1: Recording of the movement by SkillCapture or any other video capture software. In case of 3D tracking the use of two or more cameras is required.

Sequence WizardStep 2: Start SkillSpector and open the video. Major parameters is the Digitizing Model and the image calibration setup and will be setup using simple Wizards.

The Digitizing Model defines the points, objects and segments of the human body that is of interest. By having two points like ankle and knee a segment named calf (lower leg) can be defined. Points does not always have to relate to the human body but can also define other objects like golf club, ladder in high jump, football in kicking and so on.

The Calibration setup is the method by which the image is transformed into a standard coordinate system as we know it from mathematics with X, Y and Z axis with measuring units of meters or centimeters.

Video Image Digizing

Step 3: Digitizing of the movement by locating the specific reference points of the human body. SkillSpector is capable of searching for specific location in the image by the means of image processing. This means that after manually digitizing the toe the first time the software will try to relocate the new position in the next image. In some situation it is very powerful and enhance the speed of digitizing. Various settings can be adjust in order to facilitate the speed. In order to increase the accuracy of digitizing SkillSpector you can use a special SneakZoom feature that fast and easy supports subpixel accuracy. Subpixel accuracy (see picture below) is done using fast image interpolation and is not just a zoom (increase of pixel size) feature like most software I capable of today.Normal Image viewSneekZoom Image view

Digitizing 2D versus 3D: In many movement analysis the use of 2D is more quite sufficient. The decision whether to do 2D or 3D is solely a question to the nature of the motion. In this shown example (back summersault) the main movement is in one plane i.e. the sagital plane. A 2D analysis is therefor sufficient. A good thing to always keep in mind when doing motion analysis is to keep it simple in order to avoid errors and factors that don't relate to the actual movement.

Image CalibrationStep 4: Digitizing the calibration that maps the world coordinate system into image. DgeeMe is capable of using three different methods of calibration. A 2D posture calibration like the one shown in the picture on the right, a 2D calibration frame and a 3D calibration frame. The calibration frame has to be digitized in every camera view included in the sequence. Important for the calibration setup is to know the actual absolute world coordinate positions of all the calibration points. If great inaccuracy is made with the calibration points it will affect the accuracy of the actual movement measured.
Data Viewing and AnalysisStep 5: Now to the analysis and all the fun.

In the picture on the left it is the analysis of Center of gravity movement during the back summersault performed. It also includes the angle and angular velocity of the hip.

Animation RenderingThe picture on the right shows a 3D computer graphics representation of the movement. The movement can be replayed at different speeds. The orange point in the is the center of gravity.