Wageningen, 20 October 2000 - In December 1999 we released EthoVision for Windows, a major upgrade of our video tracking system for automatic measurement of movements and behavior (see Noldus News 99/2). Since then several releases with new and improved functionality have followed. Since the latest update (October 2000), the EthoVision 2.0 series is complete with a Basic, Pro and Color-Pro configuration. The color tracking algorithms of EthoVision Color-Pro have been completely redesigned, resulting in more accurate and robust color identification. EthoVision Color-Pro can now track up to 16 animals simultaneously per arena, without confusing them. Furthermore, you can now have as many as 16 arenas under a single camera (e.g. an array of 4 by 4 boxes or petri dishes). 

EthoVision Color-Pro identifies an individual by its color. If animals have natural color differences no special intervention is necessary. However, if animals are the same color, artificial color differences should be created, e.g. by marking them individually with paint, hair dye, livestock markers, etc. Prior to a trial, you 'train' the system to recognize the colors of each of the objects in the arena. During tracking, EthoVision searches each video image for the recognized colors. Groups of similarly colored pixels are classified as the objects. In practice, a tracked object can be a whole animal (for example, a yellow wasp) or a part of an animal (for example, a rat that has been marked on the head with a red dot). 

The sreenshot on the right shows the Object Identification dialog, where you tell EthoVision how to recognize the animals in the arena.

Previous versions of EthoVision Color-Pro (version 1.95 and before) only used the hue component of the hue-saturation-intensity (HSI) color space model (see figure 2). EthoVision Color-Pro 2.0 uses both the hue (color) and saturation component to track objects. Furthermore, you can set lower and upper thresholds for the intensity component. 

The interpretation of objects that are almost similar in color but differ in saturation is thus improved enormously, leading to significantly better tracking results. In addition, the use of these two complementary detection descriptors makes the object identification more robust, so that, for instance, objects can be tracked more reliably if the light intensity (brightness) is uneven across the arena. The user interface has been redesigned completely, making it much easier to train the system to identify an object by its color. In an object identification dialog box the HSI color space is represented allowing the user to interactively set the hue and saturation ranges for each of the objects. The results can be checked in the video image of the experimental setup, until all objects are detected and tracked without errors.

A color space, or color coding scheme, is a mathematical representation of a set of colors. There are many different color coding schemes, such as RGB and HSI. EthoVision Color-Pro uses the HSI color coding scheme. The HSI color coding scheme is represented by hue, saturation and intensity. It was developed to be intuitive in manipulating color, by approximating the way humans perceive and interpret color. In the HSI color space, changes in the amount of illumination across an arena (ideally) only affect the intensity component of the colors. In the double cone (see figure) you can see that the relative positions of colors are not influenced by the illumination intensity. Since both hue and saturation are used to define the objects' color, EthoVision can distinguish objects that are more similar in color to each other than if only using hue (for example objects with the same hue but differing saturation values). Due to this, EthoVision Color-Pro can now cope with as many as 16 animals per arena (under appropriate conditions). 

For more information, please contact the product manager of EthoVision.