Abstract: Disclosed in this application is an image processing method and a related apparatus. The method includes performing vectorization on a to-be-processed image, to obtain an image representation vector; performing feature mapping on the image representation vector to obtain an image feature, including: performing, at a same network layer of a network block, global feature mapping on input content to obtain a global feature, and performing local feature mapping on the input content to obtain a local feature, the input content being obtained based on the image representation vector; performing feature fusion on the global feature and the local feature to obtain a fused feature; obtaining an image feature based on the fused feature; and performing category prediction by using a classification module in the image classification model based on the image feature, to obtain a classification result of the to-be-processed image.

Abstract: Embodiments of this application provide an image data processing method, apparatus and device and a storage medium. The method includes inputting image data including text information into a text recognition model, and acquiring image representation information corresponding to the image data according to a feature extraction component in the text recognition model; obtaining semantic encoding information corresponding to the image representation information according to an image encoding component; acquiring discrete encoding information corresponding to the image representation information according to code tables included in a discrete encoding module; and correcting network parameters of the text recognition model according to an encoding similarity between the semantic encoding information and the discrete encoding information to obtain a target text recognition model.

Abstract: This disclosure provides an image processing method and processing system. The method includes obtaining an image and selecting a calibration area of the image. The method also includes reading a feature parameter corresponding to the calibration area from a preset segmentation model. The method further includes segmenting the image by using the feature parameter, to generate a segmentation result corresponding to the calibration area. This image segmentation by using the feature parameter obtained from the segmentation model provides a high accuracy segmentation rate and can be applied in wide scenarios.

Abstract: A method for detecting and tracking human head in an image by an electronic device is disclosed. The method may include segmenting the image into one or more sub-images; inputting each sub-image to a convolutional neural network trained according to training images having marked human head positions; outputting by a preprocessing layer of the convolutional neural network comprising a first convolutional layer and a pooling layer, a first feature corresponding to each sub-image; mapping through a second convolutional layer the first feature corresponding to each sub-image to a second feature corresponding to each sub-image; mapping through a regression layer the second feature corresponding to each sub-image to a human head position corresponding to each sub-image and a corresponding confidence level of the human head position; and filtering, according to the corresponding confidence level, human head positions corresponding to the one or more sub-images, to acquire detected human head positions in the image.

Abstract: The provided is an image processing method, which includes: converting a to-be-processed image into a gray image; performing Gaussian blur processing having a predefined blur radius on the gray image, and obtaining a reference image; and according to gray values of reference pixels of the reference image, decreasing a pixel value of a pixel of the to-be-processed image corresponding to a reference pixel whose gray value is larger than a gray threshold, increasing a pixel value of a pixel of the to-be-processed image corresponding to a reference pixel whose gray value is smaller than the gray threshold, and obtaining an adjusted image.

Abstract: A method for detecting and tracking human head in an image by an electronic device is disclosed. The method may include segmenting the image into one or more sub-images; inputting each sub-image to a convolutional neural network trained according to training images having marked human head positions; outputting by a preprocessing layer of the convolutional neural network comprising a first convolutional layer and a pooling layer, a first feature corresponding to each sub-image; mapping through a second convolutional layer the first feature corresponding to each sub-image to a second feature corresponding to each sub-image; mapping through a regression layer the second feature corresponding to each sub-image to a human head position corresponding to each sub-image and a corresponding confidence level of the human head position; and filtering, according to the corresponding confidence level, human head positions corresponding to the one or more sub-images, to acquire detected human head positions in the image.

Abstract: A method for detecting and tracking human head in an image by an electronic device is disclosed. The method may include segmenting the image into one or more sub-images; inputting each sub-image to a convolutional neural network trained according to training images having marked human head positions; outputting by a preprocessing layer of the convolutional neural network comprising a first convolutional layer and a pooling layer, a first feature corresponding to each sub-image; mapping through a second convolutional layer the first feature corresponding to each sub-image to a second feature corresponding to each sub-image; mapping through a regression layer the second feature corresponding to each sub-image to a human head position corresponding to each sub-image and a corresponding confidence level of the human head position; and filtering, according to the corresponding confidence level, human head positions corresponding to the one or more sub-images, to acquire detected human head positions in the image.

Abstract: This disclosure provides an image processing method and processing system. The method includes obtaining an image and selecting a calibration area of the image. The method also includes reading a feature parameter corresponding to the calibration area from a preset segmentation model. The method further includes segmenting the image by using the feature parameter, to generate a segmentation result corresponding to the calibration area. This image segmentation by using the feature parameter obtained from the segmentation model provides a high accuracy segmentation rate and can be applied in wide scenarios.

Abstract: The provided is an image processing method, which includes: converting a to-be-processed image into a gray image; performing Guass fuzzy processing having a predefined fuzzy radius on the gray image, and obtaining a reference image; and according to gray values of reference pixels of the reference image, decreasing a pixel value of a pixel of the to-be-processed image corresponding to a reference pixel whose gray value is larger than a gray threshold, increasing a pixel value of a pixel of the to-be-processed image corresponding to a reference pixel whose gray value is smaller than the gray threshold, and obtaining an adjusted image.

Abstract: The provided is an image processing method, which includes: converting a to-be-processed image into a gray image; performing Gaussian blur processing having a predefined fuzzy radius on the gray image, and obtaining a reference image; and according to gray values of reference pixels of the reference image, decreasing a pixel value of a pixel of the to-be-processed image corresponding to a reference pixel whose gray value is larger than a gray threshold, increasing a pixel value of a pixel of the to-be-processed image corresponding to a reference pixel whose gray value is smaller than the gray threshold, and obtaining an adjusted image.

Abstract: A method for implementing image denoising is provided. The method includes: calculating a tangent value of each pixel; determining whether a modulus value of the tangent value of each pixel is less than a preset threshold, if yes, determining a corresponding pixel as a non-boundary point of the image, and performing bilateral filter on a pixel determined as a non-boundary point of the image and pixels which are around the pixel and of which distances to the pixel are less than or equal to a first filtering radius; and if not, determining a corresponding pixel as a boundary point of the image, and performing bilateral filter on a pixel determined as a boundary point of the image and pixels whose distances along tangent directions and opposite directions of the tangent directions to the pixel are less than or equal to a second filtering radius.

Abstract: The provided is an image processing method, which includes: converting a to-be-processed image into a gray image; performing Guass fuzzy processing having a predefined fuzzy radius on the gray image, and obtaining a reference image; and according to gray values of reference pixels of the reference image, decreasing a pixel value of a pixel of the to-be-processed image corresponding to a reference pixel whose gray value is larger than a gray threshold, increasing a pixel value of a pixel of the to-be-processed image corresponding to a reference pixel whose gray value is smaller than the gray threshold, and obtaining an adjusted image.

Abstract: A method for implementing image denoising is provided. The method includes: calculating a tangent value of each pixel; determining whether a modulus value of the tangent value of each pixel is less than a preset threshold, if yes, determining a corresponding pixel as a non-boundary point of the image, and performing bilateral filter on a pixel determined as a non-boundary point of the image and pixels which are around the pixel and of which distances to the pixel are less than or equal to a first filtering radius; and if not, determining a corresponding pixel as a boundary point of the image, and performing bilateral filter on a pixel determined as a boundary point of the image and pixels whose distances along tangent directions and opposite directions of the tangent directions to the pixel are less than or equal to a second filtering radius.