System and method for image classification based on Tsallis entropy
AssigneeSeematics Systems Ltd
InventorMoshe Guttmann
Life of this patent
Estimating the term…
Abstract
System and method for image classification based on Tsallis entropy are provided. An image may be segmented to obtain image regions. The image regions may be analyzed to generate histograms. The histograms may be analyzed to compute Tsallis entropy values. The image regions may be associated with categories based on the Tsallis entropy values. Image regions associated with a first category may be classified to obtain classification information. The classification information may be transmitted to an external device. Information related to image regions associated with a second category may be transmitted to the external device. The external device may be configured to classify the image regions associated with the second category based on the transmitted information.
Claims17
A system comprising:
- at least one communication device;
- at least one image sensor configured to capture an image from an environment;
- and
- at least one processing unit configured to:
- segment the image to extract a plurality of image regions;
- analyze the plurality of image regions to generate a plurality of histograms associated with the plurality of image regions;
- compute a plurality of Tsallis entropy values associated with the plurality of histograms;
- based on the plurality of Tsallis entropy values, associate each image region of the plurality of image regions with a category of a plurality of categories, the plurality of categories comprises at least a first category and a second category;
- based, at least in part, on the plurality of histograms, classify image regions associated with the first category to obtain classification information;
- transmit to an external device, using the at least one communication device, the classification information;
- and
- transmit to the external device, using the at least one communication device, information related to image regions associated with the second category, wherein the external device is configured to classify the image regions associated with the second category based on the transmitted information.
A method comprising:
- capturing an image from an environment using at least one image sensor;
- segmenting the image to extract a plurality of image regions;
- analyzing the plurality of image regions to generate a plurality of histograms associated with the plurality of image regions;
- computing a plurality of Tsallis entropy values associated with the plurality of histograms;
- based on the plurality of Tsallis entropy values, associating each image region of the plurality of image regions with a category of a plurality of categories, the plurality of categories comprises at least a first category and a second category;
- based, at least in part, on the plurality of histograms, classifying image regions associated with the first category to obtain classification information;
- transmitting to an external device, using at least one communication device, the classification information;
- and
- transmitting to the external device, using the at least one communication device, information related to image regions associated with the second category, wherein the external device is configured to classify the image regions associated with the second category based on the transmitted information.
Dependencies are inferred from the claim text.