System and process for bootstrap initialization of nonparametric color models

Abstract

A system and process for automatically learning a reliable color-based tracking system is presented. The tracking system is learned by using information produced by an initial object model in combination with an initial tracking function to probabilistically determine the configuration of one or more target objects in a temporal sequence of images, and a data acquisition function for gathering observations relating to color in each image. The observations gathered by the data acquisition function include information that is relevant to parameters desired for a final color-based object model. A learning function then uses probabilistic methods to determine conditional probabilistic relationships between the observations and probabilistic target configuration information to learn a color-based object model automatically tailored to specific target objects. The learned object model is then used in combination with the final tracking function to probabilistically locate and track specific target objects in one or more sequential images.

Description

BACKGROUND 1. Technical Field The invention is related to a system and process for automatically generating a reliable color-based tracking system, and more particularly, to a system and process for using information gathered from an initial object tracking system to automatically learn a color-based object model tailored to at least one specific target object, to create a tracking system more reliable than the initial object tracking system. 2. Related Art Most current systems for determining the presence of objects of interest in an image or scene have involved processing a temporal sequence of color or grayscale images of a scene using a tracking system. Objects are typically recognized, located and / or tracked in these systems using, for example, color-based, edge-based, shape-based, or motion-based tracking schemes to process the images. While the aforementioned tracking systems are useful, they do have limitations. For example, such object tracking systems typically use a ...

AI Technical Summary

Benefits of technology

As mentioned previously, one embodiment of the present invention includes an initial contour-based tracking function for locating and tracking target objects such as human faces. This initial tracking function accepts the parameters defining an initial contour-based object model of an expected target object, such as a generic human face, in combination with one or more sequential images, and outputs a state estimate for each image. Human faces are roughly elliptical. Therefore, when tracking human faces, the initial contour-based tracking function uses adjacent frame differencing to detect moving edges in sequential images, then continues by using contour tracking to track the most salient ellipse or ellipses by comparing the detected edges to elliptical contours in the contour-based object model of a generic face. This conventional technique returns a state estimate over each image, detailing the probable configurations of one or more faces in the image. Such a technique is capable of returning a state estimate after processing a single image. However, accuracy improves with the processing of additional images.

In tracking target faces, the color-based tracking function accepts the parameters defining the learned color-based object model, in combination with one or more sequential images and outputs either a state estimate for each image, or simply target object position information with respect to each image. As with the state estimate output by the initial tracking function, the state estimate output by the color-based tracking function is a probability distribution over the entire range of the image wherein higher probabilities denote a greater likelihood of target object configuration. The color-based object model contains the information about which color ranges are specific to target objects such as faces, and which color ranges are specific to the background. Consequently, the color-based tracking function can simply examine every pixel in the image and assign it a probability, based on the measure of confidence associated with each color range, that it either belongs to the target object or to the background. Further, as discussed above, the color-based object model may be iteratively updated, thereby increasing in accuracy over time. Consequently, the accuracy of the state estimate or position information output by the color-based tracking function also increases over time as the accuracy of the color-based object model increases.

Problems solved by technology

While the aforementioned tracking systems are useful, they do have limitations.

However, it is not generally possible to create a single high fidelity object model that ideally represents each of the many potential derivatives or views of a single object type, such as the faces of different individuals having different skin coloration, facial structure, hair type and style, etc., under any of a number of lighting conditions.

Consequently, such tracking systems are prone to error, especially where the actual parameters defining the target object deviate in one or more ways from the parameters defining the generic object model.

In these cases, the value of the learned model is fleeting since few targets ever maintain constant dynamic geometries.

However, the aforementioned systems typically require manual intervention in learning or fine-tuning those tracking systems.

Consequently, it is difficult or impossible for such systems to quickly respond to the dynamic environment often associated with tracking possibly moving target objects under possibly changing lighting conditions.

Images