Detecting and labeling places using runtime change-point detection and place labeling classifiers

• US 8,559,717 B2
• Filed: 03/15/2011
• Issued: 10/15/2013
• Est. Priority Date: 03/16/2010
• Status: Active Grant

• Current Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Original Assignee: Honda Motor Co., Ltd. (Honda Motor Company)
• Inventors: Ranganathan, Ananth
• Primary Examiner(s): Zarka, David
• Assistant Examiner(s): YEH, EUENG NAN
• Application Number:
  US13/048,617
• Publication Number:
  US 20110229031A1
• Time in Patent Office: 945 Days
• Field of Search: 382/170, 382/171, 382/190, 382/225, 382/275, 348/335
• US Class Current: 382/170
• CPC Class Codes: G06V 20/10 Terrestrial scenes scenes under surveillance with static cameras G06V20/52; scenes perceived from the exterior of a vehicle G06V20/56; scenes perceived from the interior of a vehicle G06V20/59

Abstract

A system and method are disclosed for detecting and labeling places in a video stream using change-points detection. The system comprises a place label generation module configured to assign place labels probabilistically to places in the video stream based on the measurements of the measurement stream representing the video. For each measurement in the segment, the place label generation module classifies the measurement by computing the probability of the measurement being classified by a learned Gaussian Process classifier. Based on the probabilities generated with respect to all the measurements in the segment, the place label generation module determines the place label for the segment. In cases where a Gaussian Process classifier cannot positively classify a segment, the place label generation module determines whether the segment corresponds to an unknown place based on the perplexity statistics of the classification and a threshold value.

Claims

1. A computer implemented method for labeling places captured in a video stream, the method comprising:

receiving a video stream comprising multiple digital representations of images;

generating a measurement stream representing the video stream, wherein the measurement stream comprises one or more image histograms of the video stream;

segmenting the measurement stream into a plurality of segments corresponding to types of places in the video stream based on the histograms of the videos stream, wherein the boundary between two adjacent segments represents a change-point of the measurement stream at a time-step;

computing probability distributions of the segments over a plurality of place classifiers, wherein probability of a segment represents a likelihood that the segment is classified by a place classifier of the plurality of the place classifiers, computing probability distributions of the segments comprising;

projecting the measurement stream having a first dimensional space into a second dimensional space, wherein the second dimensional space has a smaller dimension than the one of the first dimensional space; and

computing predictive probability distributions of the segments in the second dimensional space; and

generating place labels for the types of places in video stream based on probabilities of change-points of the measurement stream, wherein the change-points are detected based on the probability distributions of the segments over the plurality of the place classifiers.

2. The method of claim 1, wherein generating the measurement stream comprises:

extracting image features from the images of the video stream;

for each image of the video stream;

generating one or more image histograms based on the extracted image features associated with the image, wherein each bin of an image histogram indicates the frequency of quantized image features in the image; and

generating a spatial pyramid of image histograms associated with the image, wherein the spatial pyramid of image histograms associated with the image is a measurement of the image in the measurement stream.

3. The method of claim 2, wherein generating one or more image histograms of the image of the videos stream comprises:

dividing the image into a plurality of sub-images, each sub-image has a smaller spatial resolution than the spatial resolution of the image;

for each sub-image, generating one or more image histograms based on image features associated with the sub-image.

4. The method of claim 1, wherein segmenting the measurement stream comprises:

dividing the segment stream into a plurality of non-overlapping and adjacent segments;

modeling the change-points of the measurement stream by the plurality of the place classifiers.

5. The method of claim 1, further comprising learning the plurality of place classifiers from a plurality of image representations of images of an image training set.

6. The method of claim 5, wherein learning the plurality of place classifiers comprises:

extracting image features of the images of the image training set, wherein types of places in each image of the image training set are pre-labeled;

clustering the extracted images features to create a codebook of a plurality of code words, wherein a code word is represented by a cluster identification of quantized image features associated with the code word;

generating one or more image histograms for each image of the image training set, wherein each bin of an image histogram indicates the frequency of the quantized image features in the image.

7. The method of claim 1, wherein generating place labels for the types of places in video stream comprises:

for each segment of the measurement stream, the segment comprising one or more measurements;

generating a place label for each measurement of the segment, wherein the place label has a probability indicting that the segment is classified by a place classifier of the plurality of the place classifiers;

comparing the probabilities for the one or more measurements classified by the place classifier;

selecting the place label based on the probabilities of the measurements of the segment, the selected place label positively indicated by a majority of the probabilities of the measurements of the segment.

8. The method of claim 1, wherein the plurality of place classifiers are Gaussian Process classifiers, a Gaussian Process classifier configured to positively classify a place in a video sequence to its corresponding type of place.

9. The method of claim 1, wherein generating place labels for the types of places in video stream further comprises:

responsive to a segment not recognized by the plurality of place classifiers;

determining the segment corresponding to a place unknown to the plurality of the place classifiers; and

assigning a new place label to the segment.

10. The method of claim 9, wherein determining the segment corresponding to a place unknown to the plurality of the place classifiers further comprises:

computing perplexity statistic of the classification of the segment using continuous output of the classification from the plurality of place classifiers and covariance value associated with the continuous output of the classification; and

determining whether the segment corresponds to a place unknown to the plurality of place classifiers based on the perplexity statistic of the classification.

11. The method of claim 10, wherein determining whether the segment corresponds to a place unknown to the plurality of place classifiers further comprises:

comparing the perplexity statistic of the classification with a predefined threshold value; and

determining the segment corresponds to a place unknown to the plurality of place classifiers responsive to the perplexity statistic of the classification below the threshold value.

12. A computer system for labeling places captured in a video stream, the system comprising:

a segmentation module configured to;

receive a video stream comprising multiple digital representations of images;

generate a measurement stream representing the video stream, wherein the measurement stream comprises one or more image histograms of the video stream;

segment the measurement stream into a plurality of segments corresponding to types of places in the video stream based on the histograms of the videos stream, wherein the boundary between two adjacent segments represents a change-point of the measurement stream at a time-step;

compute probability distributions of the segments over a plurality of place classifiers, wherein probability of a segment represents a likelihood that the segment is classified by a place classifier of the plurality of the place classifiers, wherein computing probability distributions of the segments comprises;

projecting the measurement stream having a first dimensional space into a second dimensional space, wherein the second dimensional space has a smaller dimension than the one of the first dimensional space; and

computing predictive probability distributions of the segments in the second dimensional space; and

a place label generation module configured to;

generate place labels for the types of places in video stream based on probabilities of change-points of the measurement stream, wherein the change-points are detected based on the probability distributions of the segments over the plurality of the place classifiers.

13. The system of claim 12, wherein the segmentation module is further configured to:

extract image features from the images of the video stream;

for each image of the video stream;

generate one or more image histograms based on the extracted image features associated with the image, wherein each bin of an image histogram indicates the frequency of quantized image features in the image; and

generate a spatial pyramid of image histograms associated with the image, wherein the spatial pyramid of image histograms associated with the image is a measurement of the image in the measurement stream.

14. The system of claim 13, wherein the segmentation module is further configured to:

divide the image into a plurality of sub-images, each sub-image has a smaller spatial resolution than the spatial resolution of the image;

for each sub-image, generate one or more image histograms based on image features associated with the sub-image.

15. The system of claim 12, wherein the segmentation module is further configured to:

divide the segment stream into a plurality of non-overlapping and adjacent segments;

model the change-points of the measurement stream by the plurality of the place classifiers.

16. The system of claim 12, further comprising a label learning module configured to learn the plurality of place classifiers from a plurality of image representations of images of an image training set.

17. The system of claim 16, further comprising:

a feature detection module configured to extract image features of the images of the image training set, wherein types of places in each image of the image training set are pre-labeled;

a feature clustering module configured to cluster the extracted images features to create a codebook of a plurality of code words, wherein a code word is represented by a cluster identification of quantized image features associated with the code word;

an image representation module configured to generate one or more image histograms for each image of the image training set, wherein each bin of an image histogram indicates the frequency of the quantized image features in the image.

18. The system of claim 12, wherein the place label generating module is further configured to:

for each segment of the measurement stream, the segment comprising one or more measurements;

generate a place label for each measurement of the segment, wherein the place label has a probability indicting that the segment is classified by a place classifier of the plurality of the place classifiers;

compare the probabilities for the one or more measurements classified by the place classifier;

select the place label based on the probabilities of the measurements of the segment, the selected place label positively indicated by a majority of the probabilities of the measurements of the segment.

19. The system of claim 12, wherein the plurality of place classifiers are Gaussian Process classifiers, a Gaussian Process classifier configured to positively classify a place in a video sequence to its corresponding type of place.

20. The system of claim 12, wherein the place label generation module is further configured to:

responsive to a segment not recognized by the plurality of place classifiers;

determine the segment corresponding to a place unknown to the plurality of the place classifiers; and

assign a new place label to the segment.

21. The system of claim 20, wherein the place label generation module is further configured to:

compute perplexity statistic of the classification of the segment using continuous output of the classification from the plurality of place classifiers and covariance value associated with the continuous output of the classification; and

determine whether the segment corresponds to a place unknown to the plurality of place classifiers based on the perplexity statistic of the classification.

22. The system of claim 21, wherein the place label generation module is further configured to:

compare the perplexity statistic of the classification with a predefined threshold value; and

determine the segment corresponds to a place unknown to the plurality of place classifiers responsive to the perplexity statistic of the classification below the threshold value.

23. A computer program product for labeling places captured in a video stream, the computer program product comprising a non-transitory computer-readable medium containing computer program code for performing the operations:

receiving a video stream comprising multiple digital representations of images;

generating a measurement stream representing the video stream, wherein the measurement stream comprises one or more image histograms of the video stream;

segmenting the measurement stream into a plurality of segments corresponding to types of places in the video stream based on the histograms of the videos stream, wherein the boundary between two adjacent segments represents a change-point of the measurement stream at a time-step;

computing probability distributions of the segments over a plurality of place classifiers, wherein probability of a segment represents a likelihood that the segment is classified by a place classifier of the plurality of the place classifiers, wherein computing probability distributions of the segments comprises;

projecting the measurement stream having a first dimensional space into a second dimensional space, wherein the second dimensional space has a smaller dimension than the one of the first dimensional space; and

computing predictive probability distributions of the segments in the second dimensional space; and

generating place labels for the types of places in video stream based on probabilities of change-points of the measurement stream, wherein the change-points are detected based on the probability distributions of the segments over the plurality of the place classifiers.

24. The computer program product of claim 23, wherein the computer program code for generating the measurement stream comprises computer program code for:

extracting image features from the images of the video stream;

for each image of the video stream;

generating one or more image histograms based on the extracted image features associated with the image, wherein each bin of an image histogram indicates the frequency of quantized image features in the image; and

generating a spatial pyramid of image histograms associated with the image, wherein the spatial pyramid of image histograms associated with the image is a measurement of the image in the measurement stream.

25. The computer program product of claim 24, wherein the computer program code for generating one or more image histograms of the image of the videos stream comprises computer program code for:

dividing the image into a plurality of sub-images, each sub-image has a smaller spatial resolution than the spatial resolution of the image;

for each sub-image, generating one or more image histograms based on image features associated with the sub-image.

26. The computer program product of claim 23, wherein the computer program code for segmenting the measurement stream comprises computer program code for:

dividing the segment stream into a plurality of non-overlapping and adjacent segments;

modeling the change-points of the measurement stream by the plurality of the place classifiers.

27. The computer program product of claim 23, further comprising computer program code for learning the plurality of place classifiers from a plurality of image representations of images of an image training set.

28. The computer program product of claim 27, wherein the computer program code for learning the plurality of place classifiers comprises computer program code for:

extracting image features of the images of the image training set, wherein types of places in each image of the image training set are pre-labeled;

clustering the extracted images features to create a codebook of a plurality of code words, wherein a code word is represented by a cluster identification of quantized image features associated with the code word;

generating one or more image histograms for each image of the image training set, wherein each bin of an image histogram indicates the frequency of the quantized image features in the image.

29. The computer program product of claim 23, wherein computer program code for generating place labels for the types of places in video stream comprises computer program code for:

for each segment of the measurement stream, the segment comprising one or more measurements;

generating a place label for each measurement of the segment, wherein the place label has a probability indicting that the segment is classified by a place classifier of the plurality of the place classifiers;

comparing the probabilities for the one or more measurements classified by the place classifier;

selecting the place label based on the probabilities of the measurements of the segment, the selected place label positively indicated by a majority of the probabilities of the measurements of the segment.

30. The computer program product of claim 23, wherein the plurality of place classifiers are Gaussian Process classifiers, a Gaussian Process classifier configured to positively classify a place in a video sequence to its corresponding type of place.

31. The computer program product of claim 23, wherein computer program code for generating place labels for the types of places in video stream further comprises computer program code for:

responsive to a segment not recognized by the plurality of place classifiers;

determining the segment corresponding to a place unknown to the plurality of the place classifiers; and

assigning a new place label to the segment.

32. The computer program product of claim 31, wherein computer program code for determining the segment corresponding to a place unknown of the place classifiers further comprises computer program code for:

computing perplexity statistic of the classification of the segment using continuous output of the classification from the plurality of place classifiers and covariance value associated with the continuous output of the classification; and

determining whether the segment corresponds to a place unknown to the plurality of place classifiers based on the perplexity statistic of the classification.