Edge detection using fuzzy gradient information
AssigneeAdobe Inc.
InventorSomya Jain
Life of this patent
Estimating the term…
Abstract
Techniques of detecting edges of objects in images include applying an adaptive threshold to a sigmoidal membership function of a fuzzy gradient of the image brightness. For example, when a computer configured to detect edges of an object of an image receives a grayscale image, the computer computes gradients of the image brightness at each pixel to form a gradient matrix. The computer then creates a fuzzy gradient matrix by applying a sigmoidal membership function to each element of the gradient matrix to form a smoothed gradient matrix. The computer then performs an adaptive threshold operation on the smoothed gradient matrix to produce the edges of the object of the image.
Claims20
A computer-implemented method for producing edges of an object in an image, comprising:
- receiving image data representing a plurality of pixels of the image, each of the plurality of pixels including a brightness value;
- generating gradient matrix data representing a gradient matrix having elements, each element of the gradient matrix corresponding to a respective pixel of the plurality of pixels and including a gradient of the brightness value of the pixel;
- generating membership function data representing a membership function matrix having elements, each element of the membership function matrix including a sigmoidal function of a respective element of the gradient matrix;
- and
- performing an adaptive threshold operation on the membership function data to produce the edges of the object in the image, the adaptive threshold operation including generating a local threshold corresponding to a subset of the plurality of pixels of the image, the local threshold being based on the brightness values of the pixels of the subset of the plurality of pixels of the image.
A computer program product comprising a nontransitory storage medium, the computer program product including code that, when executed by processing circuitry of a computer configured to produce edges of an object in an image, cause the computer to perform a method, the method comprising:
- receiving image data representing a plurality of pixels of the image, each of the plurality of pixels including a brightness value;
- generating gradient matrix data representing a gradient matrix having elements, each element of the gradient matrix corresponding to a respective pixel of the plurality of pixels and including a gradient of the brightness value of the pixel;
- generating membership function data representing a membership function matrix having elements, each element of the membership function matrix including a sigmoidal function of a respective element of the gradient matrix;
- and
- performing an adaptive threshold operation on the membership function data to produce the edges of the object in the image, the adaptive threshold operation including generating a local threshold corresponding to a subset of the plurality of pixels of the image, the local threshold being based on the brightness values of the pixels of the subset of the plurality of pixels of the image.
An electronic apparatus configured to produce edges of an object in an image, the electronic apparatus comprising:
- a network interface;
- a memory;
- and
- controlling circuitry coupled to the memory, the controlling circuitry being configured to:
- receive image data representing a plurality of pixels of the image, each of the plurality of pixels including a brightness value;
- generate gradient matrix data representing a gradient matrix having elements, each element of the gradient matrix corresponding to a respective pixel of the plurality of pixels and including a gradient of the brightness value of the pixel;
- generate membership function data representing a membership function matrix having elements, each element of the membership function matrix including a sigmoidal function of a respective element of the gradient matrix;
- and
- perform an adaptive threshold operation on the membership function data to produce the edges of the object in the image, the adaptive threshold operation including generating a local threshold corresponding to a subset of the plurality of pixels of the image, the local threshold being based on the brightness values of the pixels of the subset of the plurality of pixels of the image.
Dependencies are inferred from the claim text.