METHOD AND APPARATUS FOR OBTAINING AND PROCESSING IMAGE FEATURES

US 20090285473A1
Filed: 12/20/2007
Published: 11/19/2009
Est. Priority Date: 12/20/2007
Status: Active Grant

First Claim

Patent Images

1. A method, comprising:

training groups of training features derived from regions of training images to obtain a plurality of classifiers, wherein each classifier corresponds to each group of training features,utilizing the plurality of classifiers to classify groups of validation features derived from regions of validation images to obtain a plurality of weights, wherein each weight corresponds to each region of the validation images and indicates how important the each region of the validation images is; and

discarding a weight from the plurality of weights based upon a certain criterion.

View all claims

2 Assignments

Timeline View

Assignment View

0 Petitions

Accused Products

Abstract

Machine-readable media, methods, apparatus and system for obtaining and processing image features are described. In some embodiments, groups of training features derived from regions of training images may be trained to obtain a plurality of classifiers, each classifier corresponding to each group of training features. The plurality of classifiers may be used to classify groups of validation features derived from regions of validation images to obtain a plurality of weights, wherein each weight corresponds to each region of the validation images and indicates how important the each region of the validation images is. Then, a weight may be discarded from the plurality of weights based upon a certain criterion.

Citations

19 Claims

1. A method, comprising:
- training groups of training features derived from regions of training images to obtain a plurality of classifiers, wherein each classifier corresponds to each group of training features,utilizing the plurality of classifiers to classify groups of validation features derived from regions of validation images to obtain a plurality of weights, wherein each weight corresponds to each region of the validation images and indicates how important the each region of the validation images is; and
  
  discarding a weight from the plurality of weights based upon a certain criterion.
- View Dependent Claims (2, 3, 4, 5, 6)
- - 2. The method of claim 1, wherein the weight is a negative weight.
  - 3. The method of claim 1, further comprising:
    - obtaining the groups of training features from the regions of the training images, wherein the each group of training features corresponds to each region of the training images; and
      
      obtaining the groups of validation features from the regions of the validation images, wherein each group of validation features corresponds to the each region of the validation images and the each region of the validation images corresponds to the each region of the training images.
  - 4. The method of claim 1, wherein the utilizing the plurality of classifiers further comprises:
    - utilizing the plurality of classifiers to classify the groups of validation features to obtain a plurality of similarity factors, wherein each similarity factor corresponds to the each region of the validation images and indicates how close the each region of the validation images is to a corresponding region of the training images; and
      
      obtaining the plurality of weights based upon the plurality of similarity factors.
  - 5. The method of claim 1, further comprising:
    - utilizing a remaining weight from the plurality of weights to optimize the plurality of classifiers, to provide an optimized classifier;
      
      determining a region from a test image with reference to the remaining weight;
      
      obtaining a group of test features from the region of the test image; and
      
      utilizing the optimized classifier to classify the group of test features.
  - 6. The method of claim 1, wherein the obtaining the plurality of weights further comprising:
    - establishing a discriminant function $η (z) = \log \frac{P (t = 1  z)}{P (t = 0  z)} = \log \frac{P (t = 1  z)}{1 - P (t = 1  z)} = β^{T} z,$ wherein P represents a possibility, t=1 represents a validation image of the validation images positively matches to a training image of the training images, t=0 represents the validation image negatively matches to the training image, z represents how close the each region of the validation image is to the each region of the training image, β
      
      ^Trepresents the plurality of weights;
      
      establishing a likelihood function $L (z) = \sum_{t (z_{i}) = 1} η (z_{i}) - \sum_{t (z_{j}) = 0} η (z_{j}),$ wherein $\sum_{t (z_{i}) = 1} η (z_{i})$ represents a sum of discriminant functions for one or more of the validation images that positively match to one or more of the training images, and $\sum_{t (z_{j}) = 0} η (z_{j})$ represents a sum of discriminant functions for one or more of the validation images that negatively match to one or more of the training images; and
      
      obtaining the plurality of weights by maximizing the likelihood function.

7. An apparatus, comprising:
- a classifier logic to train groups of training features derived from regions of training images to obtain a plurality of classifiers, wherein each classifier corresponds to each group of training features;
  
  a classifying logic to utilize the plurality of classifiers to classify groups of validation features derived from regions of validation images to obtain a plurality of similarity factors, wherein each similarity factor corresponds to each region of the validation images;
  
  a weight logic to obtain a plurality of weights based upon the plurality of similarity factors, wherein each weight corresponds to the each region of the validation images and indicates how important the each region of the validation images is; and
  
  a weight optimizing logic to optimize the plurality of weights by discarding a weight from the plurality of weights based upon a certain criterion.
- View Dependent Claims (8, 9, 10, 11, 12, 13)
- - 8. The apparatus of claim 7, wherein the weight is a negative weight.
  - 9. The apparatus of claim 7, further comprising a feature logic to obtain the groups of training features from the regions of the training images, wherein the each group of training features corresponds to each region of the training images;
    - and to obtain the groups of validation features from the regions of the validation images, wherein each group of validation features corresponds to the each region of the validation images and the each region of the validation images corresponds to the each region of the training images.
  - 10. The apparatus of claim 7, wherein the each similarity factor indicates how close the each region of the validation images is to a corresponding region of the training images.
  - 11. The apparatus of claim 7, further comprising a classifier optimizing logic to utilize a remaining weight from the plurality of weights to optimize the plurality of classifiers to provide an optimized classifier.
  - 12. The apparatus of claim 11, wherein the feature logic is further to determine a region from a test image with reference to the remaining weight and obtain a group of test features from the region of the test image;
    - and the classifying logic is further to utilize the optimized classifier to classify the group of test features.
  - 13. The apparatus of claim 7, wherein the weight logic is further to:
    - establishing a discriminant function $η (z) = \log \frac{P (t = 1  z)}{P (t = 0  z)} = \log \frac{P (t = 1  z)}{1 - P (t = 1  z)} = β^{T} z,$ wherein P represents a possibility, t=1 represents a validation image of the validation images positively matches to a training image of the training images, t=0 represents the validation image negatively matches to the training image, z represents how close each region of the validation image is to each region of the training image, β
      
      ^Trepresents the plurality of weights;
      
      establishing a likelihood function $L (z) = \sum_{t (z_{i}) = 1} η (z_{i}) - \sum_{t (z_{j}) = 0} η (z_{j}),$ wherein $\sum_{t (z_{i}) = 1} η (z_{i})$ represents a sum of discriminant functions for one or more of the validation images that positively match to one or more of the training images, and $\sum_{t (z_{j}) = 0} η (z_{j})$ represents a sum of discriminant functions for one or more of the validation images that negatively match to one or more of the training images; and
      
      obtaining the plurality of weights by maximizing the likelihood function.

14. A system, comprising:
- a processor;
  
  a machine-readable medium coupled to the processor, wherein the machine-readable medium further comprises a plurality of instructions which, if executed by the processor, cause the system to perform operations comprising;
  
  training groups of training features derived from regions of training images to obtain a plurality of classifiers, wherein each classifier corresponds to each group of training features,utilizing the plurality of classifiers to classify groups of validation features derived from regions of validation images to obtain a plurality of weights, wherein each weight corresponds to each region of the validation images and indicates how important the each region of the validation images is; and
  
  discarding a weight from the plurality of weights based upon a certain criterion.
- View Dependent Claims (15, 16, 17, 18, 19)
- - 15. The system of claim 14, wherein the weight is a negative weight.
  - 16. The system of claim 14, wherein the plurality of instructions further cause the system to perform operations comprising:
    - obtaining the groups of training features from the regions of the training images, wherein the each group of training features corresponds to each region of the training images; and
      
      obtaining the groups of validation features from the regions of the validation images, wherein each group of validation features corresponds to the each region of the validation images and the each region of the validation images corresponds to the each region of the training images.
  - 17. The system of claim 14, wherein the plurality of instructions that cause the system utilizing the plurality of classifiers, further cause the system to perform operations comprising:
    - utilizing the plurality of classifiers to classify the groups of validation features to obtain a plurality of similarity factors, wherein each similarity factor corresponds to the each region of the validation images and indicates how close the each region of the validation images is to a corresponding region of the training images; and
      
      obtaining the plurality of weights based upon the plurality of similarity factors.
  - 18. The system of claim 14, the plurality of instructions further cause the system to perform operations comprising:
    - utilizing a remaining weight from the plurality of weights to optimize the plurality of classifiers, to provide an optimized classifier;
      
      determining a region from a test image with reference to the remaining weight;
      
      obtaining a group of test features from the region of a test image; and
      
      utilizing the optimized classifier to classify the group of test features.
  - 19. The system of claim 14, wherein the plurality of instructions that cause the system obtaining the plurality of weights, further cause the system to perform operations comprising:
    - establishing a discriminant function $η (z) = \log \frac{P (t = 1  z)}{P (t = 0  z)} = \log \frac{P (t = 1  z)}{1 - P (t = 1  z)} = β^{T} z,$ wherein P represents a possibility, t=1 represents a validation image of the validation images positively matches to a training image of the training images, t=0 represents the validation image negatively matches to the training image, z represents how close the each region of the validation image is to the each region of the training image, β
      
      ^Trepresents the plurality of weights;
      
      establishing a likelihood function $L (z) = \sum_{t (z_{i}) = 1} η (z_{i}) - \sum_{t (z_{j}) = 0} η (z_{j}),$ wherein $\sum_{t (z_{i}) = 1} η (z_{i})$ represents a sum of discriminant functions for one or more of the validation images that positively match to one or more of the training images, and $\sum_{t (z_{j}) = 0} η (z_{j})$ represents a sum of discriminant functions for one or more of the validation images that negatively match to one or more of the training images; and
      
      obtaining the plurality of weights by maximizing the likelihood function.

Specification

Resources

Litigation Campaign Assessment

Current Assignee
Beijing Xiaomi Mobile Software Co., Ltd. (Xiaomi Corp.)
Original Assignee
Intel Corporation
Inventors
Wang, Tao, Li, Jianguo, Zhang, Yimin

Granted Patent

US 8,131,065 B2
Time in Patent Office

Days
Field of Search
US Class Current

382/159
CPC Class Codes

G06F 18/214   Generating training pattern...

G06V 10/449   Biologically inspired filte...

G06V 10/774   Generating sets of training...

METHOD AND APPARATUS FOR OBTAINING AND PROCESSING IMAGE FEATURES

First Claim

2 Assignments

0 Petitions

Accused Products

Abstract

Citations

19 Claims

Specification

Solutions

Use Cases

Quick Links

METHOD AND APPARATUS FOR OBTAINING AND PROCESSING IMAGE FEATURES

First Claim

2 Assignments

Subscription Required

Subscription Required

0 Petitions

Subscription Required

Accused Products

Subscription Required

Abstract

Citations

19 Claims

Specification

Subscription Required

Solutions

Use Cases

Quick Links