Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A method for encoding image pixels and a method for processing images

A pixel, image technology applied in the field of image recognition in which objects are reproduced to achieve the effect of increasing knowledge and expertise, avoiding errors in image recognition and analysis, and improving reliability and accuracy

Inactive Publication Date: 2009-03-18
BRACCO IMAGINIG SPA +1
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Currently there are no rules defining such relationships other than those that are defined on the basis of assumptions or speculatively based on specific properties of objects represented by images

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A method for encoding image pixels and a method for processing images
  • A method for encoding image pixels and a method for processing images
  • A method for encoding image pixels and a method for processing images

Examples

Experimental program
Comparison scheme
Effect test

example 1

[0147] Example 1 (Figure 5)

[0148] object Chest imaging method nuclear magnetic resonance Purpose identify tissue type organization type

1. Benign tumor

2. Malignant tumor pixel encoding

3×3 pixel matrix

Encode the pixel as the center pixel Image Resolution 256×256 pixels

[0149] In Example 1, a training database for an image processing algorithm is generated that recognizes two types of tissue, benign and malignant tissue in the thoracic region.

[0150] A predetermined number of MRI images of chest regions of patients diagnosed with malignant breast tumors and patients with benign breast tumors were pixel-coded according to the method described above. The determination vector of pixels has as components all the pixels of the surrounding 3x3 pixel matrix, where the pixel being coded is the center pixel (Fig. 1).

[0151] Assign the tissue type reproduced by the pixel in the image to each pixel's identified vecto...

example 2

[0170] Example 2 is similar to Example 1, but a tissue type, normal tissue, is added to the recognition database. Thus, when generating the training database, the encoded vectors of image pixels are uniquely associated with one of the tissue types represented therein, benign, malignant, or normal.

[0171] This additional type allows a large number of pixels and corresponding definite vectors to have definite meaning. In the previous example, these pixel determination vectors and the corresponding pixels have no meaning in the processing algorithm, whereas in the second example the processing algorithm can assign an additional unambiguous class or tissue type.

[0172] The additional possibility reduces error because the algorithm can choose from three conditions or states assigned to a certain vector.

[0173] The execution of this example can be described and shown with reference to the previous Example 1.

[0174] In this example, only the following neural network was sel...

example 3

[0179] Example 3 is similar to the above example, but includes 5 tissue types, namely: benign tumor tissue, malignant tumor tissue, normal tissue, muscle tissue and image background.

[0180] The training database is generated according to the method described above with reference to the previous example, and includes pixel determination vectors, wherein each pixel determination vector is uniquely assigned one of the above five types, that is, the type represented by a single pixel.

[0181] Figure 7 illustrates the results of tissue type recognition for each image pixel, the algorithm used is as follows:

[0182] processing algorithm Metanet Af MetaBayes MetanetBp MetanetCm FF-Bm FF-Sn FF-BP FF-Cm LDA

[0183] All of these processing algorithms except LDA are neural networks. LDA is a discriminative algorithm.

[0184] The obtained results showed a very high reliability of tissue type identification. It is further ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

A method of encoding digital or digitized image pixels. The image consists of a group of image points called pixels in two-dimensional images and three-dimensional pixels in three-dimensional. Each pixel or three-dimensional pixel is composed of corresponding display A set of values ​​representing the visual appearance of a pixel on a screen or printed image. According to the invention, at least a portion of the pixels or voxels of interest in a digital or digitized image, or each pixel or voxel in a group of pixels or voxels making up the image, is uniquely identified by a vector whose components are to be encoded by The data of the pixel or voxel and the data of at least one or some or all of the pixels surrounding the pixel to be encoded are given, and arranged as pixels or voxels included in the entire group of pixels or voxels constituting the image A predetermined subset of .

Description

technical field [0001] The invention relates above all to a method of encoding digital or digitized image pixels, which processes the information content of each pixel so that it can be used in automatic image processing systems, especially designed for image recognition of objects reproduced therein. Background technique [0002] The term "digital" is used to define images acquired by imaging devices that output images in digital format, ie, digital cameras, magnetic resonance imaging devices, ultrasound imaging devices, and other imaging devices. [0003] The term "digitized image" relates to an image that is essentially acquired and provided by an analog device that is further scanned by a device called a scanner, regardless of whether the latter is a hardware device, such as reading a physical print on a medium A device for an image, or a software device designed to sample an image and provide a collection of signals and convert it to a digital signal. [0004] The incr...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06F19/00G06T5/20A61B6/00A61B5/055A61B8/00G06T1/00G06T7/00G06T9/20
CPCG06T5/20G06T7/0012
Inventor 保罗·马西莫·布谢马
Owner BRACCO IMAGINIG SPA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com