Classification of objects through model ensembles

• US 6,778,705 B2
• Filed: 02/27/2001
• Issued: 08/17/2004
• Est. Priority Date: 02/27/2001
• Status: Expired due to Fees

• Current Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Original Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Inventors: Philomin, Vasanth, Gutta, Srinivas
• Primary Examiner(s): Mariam, Daniel
• Application Number:
  US09/794,443
• Publication Number:
  US 20020181785A1
• Time in Patent Office: 1,267 Days
• Field of Search: 382/103, 382/159, 382/173, 382/224, 382/236, 382/190, 382/261-264, 707/1, 707/7
• US Class Current: 382/224
• CPC Class Codes: G06F 18/2414 Smoothing the distance, e.g. radial basis function networks [RBFN]
• CPC Class Codes: G06T 7/00 Image analysis
• CPC Class Codes: G06V 10/764 using classification, e.g. of video objects

Abstract

A method for classification of objects in video image data. The method including the steps of: detecting moving objects in the image data; extracting two or more features from each detected moving object in the image data; classifying each moving object for each of the two or more features according to a classification method; and deriving a classification for each moving object based on the classification method for each of the two or more features. Also provided is an apparatus for classification of objects in video image data.

Claims

1. A method for classification of objects in video image data, the method comprising the steps of:

detecting moving objects in the image data;

smoothing the image data to reduce the effects of noise and then applying a derivative operator over the image data;

extracting two or more features from each detected moving object in the image data;

classifying each moving object for each of the two or more features according to a classification method, wherein the classifying comprises using at least two different classification methods for at least two of the two or more features; and

deriving a classification for each moving object based on the classification method for each of the two or more features.

2. The method of claim 1, wherein the detecting step comprises subtracting a background scene from the image data.

3. The method of claim 1, further comprising the step of filtering out unintended moving objects from the detected moving objects.

4. The method of claim 3, wherein the filtering step comprises filtering out unintended moving objects according to a detected speed and aspect ratio of each detected moving object.

5. The method of claim 1, wherein the extracting step comprises extracting at least two of x-gradient, y-gradient, and combined xy-gradient features from each of the detected moving objects.

6. The method of claim 1, wherein the classifying step comprises a Radial Basis Function Network for training and classifying at least one of the detected moving objects.

7. The method of claim 6, wherein the training step comprises:

(a) initializing the Radial Basis Function Network, the initializing step comprising the steps of;

fixing the network structure by selecting a number of basis functions F, where each basis function I has the output of a Gaussian non-linearity;

determining the basis function means μ

_I, where I=1, . . . , F, using a K-means clustering algorithm;

determining the basis function variances σ

_I²; and

determining a global proportionality factor H, for the basis function variances by empirical search;

(b) presenting the training, the presenting step comprising the steps of;

inputting training patterns X(p) and their class labels C(p) to the classification method, where the pattern index is p=1, . . . , N;

computing the output of the basis function nodes y_I(p), F, resulting from pattern X(p);

computing the F×

F correlation matrix R of the basis function outputs; and

computing the F×

M output matrix B, where d_j is the desired output and M is the number of output classes and j=1, . . . , M; and

(c) determining weights, the determining step comprising the steps of;

inverting the F×

F correlation matrix R to get R<sup>−

1</sup>; and

solving for the weights in the network.

8. The method of claim 7, wherein the classifying step comprises:

presenting each of the two or more features X_test for each detected moving object to the classification method; and

classifying each X_test by;

computing the basis function outputs, for all F basis functions;

computing output node activations; and

selecting the output z_j with the largest value and classifying X_test as a class j.

9. The method of claim 1, wherein the classifying step comprises outputting a class label identifying a class to which the detected moving object corresponds to and a probability value indicating the probability with which the unknown pattern belongs to the class for each of the two or more features.

10. The method of claim 9, wherein the deriving step comprises averaging the probability values for the two or more features for each detected moving object and determining if the average is greater than a threshold value.

11. The method of claim 9, wherein the deriving step comprises determining if there exists a majority of class labels which identify a like class.

12. An apparatus for classification of objects in video image data, the apparatus comprising:

means for detecting moving objects in the image data;

means for smoothing the image data to reduce the effects of noise;

means for applying a derivative operator over the image data;

means for extracting two or more features from each detected moving object in the image data;

means for classifying each moving object for each of the two or more features according to a classification method, wherein the means for classifying comprises means for using at least two different classification methods for at least two of the two or more features; and

means for deriving a classification for each moving object based on the classification method for each of the two or more features.

13. The apparatus of claim 12, further comprising means for filtering out unintended moving objects from the detected moving objects.

14. The apparatus of claim 12, wherein the means for extracting extracts at least two of x-gradient, y-gradient, and combined xy-gradient features from each of the detected moving objects and the apparatus further comprises means for smoothing the image data to reduce the effects of noise and applying a derivative operator over the image data prior to the extracting.

15. A program storage device readable by machine, tangibly embodying a program of instructions executable by the machine to perform method steps for classification of objects in video image data, the method comprising the steps of:

detecting moving objects in the image data;

smoothing the image data to reduce the effects of noise and then applying a derivative operator over the image data;

extracting two or more features from each detected moving object in the image data;

classifying each moving object for each of the two or more features according to a classification method, wherein classifying comprises using at least two different classification methods for at least two of the two or more features; and

deriving a classification for each moving object based on the classification method for each of the two or more features.

16. A computer program product embodied in a computer-readable medium for classification of objects in video image data, the computer program product comprising:

computer readable program code means for detecting moving objects in the image data;

computer readable program code for smoothing the image data to reduce the effects of noise and then applying a derivative operator over the image data;

computer readable program code means for extracting two or more features from each detected moving object in the image data;

computer readable program code means for classifying each moving object for each of the two or more features according to a classification method, wherein the classifying comprises using at least two different classification methods for at least two of the two or more features; and

computer readable program code means for deriving a classification for each moving object based on the classification method for each of the two or more features.