Method for detecting weak edge in noise image
A technique for detecting noise, in images, applied in the field of image edge detection, which can solve problems such as long running time, low efficiency, and reduction of weak edge contrast
Inactive Publication Date: 2017-11-17
SHENZHEN WEITESHI TECH
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] Aiming at the problems of reducing the contrast of weak edges, long running time, low efficiency, etc., the purpose of the present invention is to provide a method for detecting weak edges in noisy images. First, the point spread function is used for sampling, the image pixels are modeled, and then the Each candidate curve is searched for a sufficiently strong matched filter response, followed by detection of all actual edges in the original image, and finally a threshold is derived to detect the faintest actual edges, extrapolating the minimum detectable contrast to determine whether very detectable edges can be fully detected faint edge
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0057] It should be noted that the embodiments in the present application and the features of the embodiments may be combined with each other without conflict, and the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0058] figure 1 It is a system flow chart of a method for detecting weak edges in a noise image according to the present invention. Mainly include problem setting, search curve, edge detection, detection threshold and minimum detection contrast.
[0059] problem setting, let I 0 (x, y) represents a rectangular domain continuous noise-free function on ; the observed image is I 0 discrete and noisy versions of , of size N=n 1 ×n 2 pixel; ideally sampled with the delta point spread function for simplicity, the observed image pixel can be modeled as I i，j =I 0 (x i , y i )+ξ i，j , where ξ is additive noise; given a noisy image I, edge detection is to detect the original image...
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
Login to View More Abstract
The invention provides a method for detecting a weak edge in a noise image. The main content of the method includes: problem setting, curve search, edge detection, and detection of a threshold and the minimum detection contrast. The process includes: firstly performing sampling by employing a point spread function, and performing modeling on image pixels; then searching each candidate curve to obtain enough strong response of a matched filter; detecting all practical edges in an original image; and finally exporting a threshold, detecting the weakest practical edge, and deducting the minimum detectable contrast to determine whether the very weak edge can be completely detected. According to the method, the operation time is short, the efficiency is greatly improved, edge smoothness is avoided, the edge positioning is greatly improved, a lot of candidate edges can be effectively searched, and the error rate is reduced; and noise is reduced by employing a Gaussian filter, but the influence of change of the contrast between the edges can be avoided.
Description
technical field [0001] The invention relates to the field of image edge detection, in particular to a method for detecting weak edges in noise images. Background technique [0002] Edge detection technology is an important part of image analysis and an important preprocessing technology in the field of image processing, so it has received extensive attention and attention. Due to the advantages and disadvantages of image quality in practical applications and the complexity of the target, edge detection often needs to go through processes such as image enhancement and denoising. With the development of basic theories such as signal processing, fuzzy mathematics, and geometry, more and more More and more new technologies are introduced into edge detection methods, such as wavelet, gray correlation degree, gray prediction model, fuzzy, genetic algorithm neural network, morphology, fractal theory and other technologies. Edge detection can be applied to license plate recognition...
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
Login to View More IPC IPC(8): G06T7/136G06T7/13
CPCG06T7/13G06T7/136
Inventor 夏春秋
Owner SHENZHEN WEITESHI TECH