scispace - formally typeset
Search or ask a question
Patent

Local three-dimensional maximum inter-class variance segmentation method of three-dimensional CT image

TL;DR: In this paper, a local three-dimensional maximum inter-class variance segmentation method of a 3D CT image is presented, which belongs to the field of image segmentation, in order to solve difficult segmentation and mis-segmentation problems caused by a non-uniform gray phenomenon of a CT image.
Abstract: The present invention discloses a local three-dimensional maximum inter-class variance segmentation method of a three-dimensional CT image, and belongs to the field of image segmentation, in order tosolve difficult segmentation and mis-segmentation problems caused by a non-uniform gray phenomenon of a CT image. The method mainly comprises the following steps: (1) initializing the size of the statistics calculation window, the weight template and the size of the three-dimensional local space, and carrying out continuation processing on the original image; (2) using the one-dimensional maximuminter-class variance segmentation algorithm and the edge tracking algorithm to extract the target contour of each layer of the image, and then using the morphological dilation method to obtain the marker image of the edge transition region; and (3) according to the pixel gray level, the weighted average of the neighborhood, and the weighted median of the neighborhood, calculating the three-dimensional histogram of each three-dimensional local space, and constructing a lookup table by using a recursive method to obtain the optimal segmentation threshold. Compared with the prior art, the methodprovided by the present invention has advantages that a three-dimensional CT image with uneven grayscale can be segmented and the time complexity is O(L3).
Citations
More filters
Patent
12 Mar 2019
TL;DR: Li et al. as mentioned in this paper proposed a self-adaptive fast segmentation algorithm based on two-dimensional Otsu, which divides an image to be segmented into a target region and a background region, and sets an adaptive threshold search interval.
Abstract: The invention discloses a rail defect segmentation self-adaptive fast algorithm based on two-dimensional Otsu, which comprises establishing a two-dimensional Otsu segmentation direct division map, dividing an image to be segmented into a target region and a background region, and setting an adaptive threshold search interval. A pixel point of a threshold retrieval area is partitioned, for the graylevel of each region, Compare the mean variance of each region, find out the region where the best threshold point is, then calculate the inter-class variance of each pixel gray value in this region,and get the maximum inter-class variance value, the corresponding gray value is the best segmentation threshold value, according to the best segmentation threshold value, the image is binarized intothe target region and the background region. When the rail defect image is segmented, the threshold search interval is adaptively set to reduce redundancy calculation in the threshold search process,meanwhile, the threshold interval narrows the gap between the background and the target area, reduces the interference caused by the noise point to the threshold determination, and solves the problemthat the defect in the small area is difficult to be segmented.

2 citations

Patent
22 Oct 2019
TL;DR: Zhang et al. as mentioned in this paper presented a target detection method based on the combination of a depth map slice and a neural network, which consists of collecting an RGB-D image containing a target object, slicing a depth maps contained in the RGB-d image, superposing the sliced images containing the contour of the target object to obtain a target contour image, and inputting the target contours map after being subjected to slicing treatment and the RGB image after being transformed into a deep convolutional neural network to obtain the target detection segmentation result.
Abstract: The embodiment of the invention discloses a target detection method and a target detection system based on the combination of a depth map slice and a neural network. The method comprises the followingsteps: collecting an RGB-D image containing a target object; slicing a depth map contained in the RGB-D image, superposing the sliced images containing the contour of the target object to obtain a target contour image; carrying out convolution treatment on an RGB image in the RGB-D image; inputting the target contour map after being subjected to slicing treatment and the RGB image after being subjected to convolution treatment into a deep convolutional neural network to obtain a target detection segmentation result. According to the method, the target shape in the scene can be effectively segmented by adopting a depth map segmentation method, and the segmented shape is used as the input of the target detection neural network, so that the defect that the neural network only pays attentionto detail features and ignores contour features is powerfully overcome, and the stability and the anti-interference capability of target detection are greatly improved.
References
More filters
Patent
24 May 2017
TL;DR: In this article, a two-dimensional maximum between-class variance threshold value method is proposed to reduce the search space of a solution of the variance through the gray average value and the standard deviation, traversing search space, and recording a solution, which enables the between class variance to be the maximum, to be an optimal one-dimensional threshold value T0.
Abstract: The invention relates to a fast noise-containing image two-dimensional maximum between-class variance threshold value method, which comprises the steps of firstly solving a gray average value and a gray standard deviation of a noise image; smoothing each pixel of the image by adopting an average gray value of a 3*3 neighborhood to acquire a smooth image; then calculating the between-class variance of the smooth image by using a maximum between-class variance threshold value method, reducing the search space of a solution of the between-class variance through the gray average value and the standard deviation, traversing the search space, and recording a solution, which enables the between-class variance to be the maximum, to be an optimal one-dimensional threshold value T0; and calculating a trace of a between-class variance dispersion matrix of a target class and a background class by using a two-dimensional maximum between-class variance method, reducing the search space of a solution of the trace through the optimal one-dimensional threshold value T0 and the gray standard deviation of the noise image, traversing the search space of the solution, and recording a gray value binary group, which enables the trace of the dispersion matrix to be the maximum, to be an optimal two-dimensional cutting threshold value. The method provided by the invention can avoid traversal for all gray levels, and also can acquire an accurate solution while greatly reducing the calculation amount.

7 citations

Patent
01 Jun 2011
TL;DR: In this paper, a splitting method of a CD image of a solid engine was proposed to solve the problem of low efficiency and poor accuracy in the CD detection image of the solid engine by the manual interpretation.
Abstract: The invention discloses a splitting method of a CD image of a solid engine. The method comprises the following steps: a CD faultage image data of the solid engine is read in, an artifact which is formed outside the solid engine owning to the air agitation formed in the CD detection is eliminated; the image, the artifact of which is eliminated, filters and eliminates background noise by median filter, meanwhile, the edges of each component of the solid engine are sharpened in the CD image and a casing, a propelling agent and an astropyle of the solid engine is cut in the CD image; a defect is cut from the propelling agent. By adopting the method, an outer layer air ring artifact, the casing, the propelling agent, the astropyle and various defects of the solid engine in the CD image are cutautomatically and accurately, thus solving the problem of low efficiency and poor accuracy in the CD detection image of the solid engine by the manual interpretation as well as being not limited by the eigenvalue of image information.

5 citations

Patent
16 Sep 2015
TL;DR: In this article, a CT image-based three-dimensional segmentation method was proposed to reduce the dependence on experience of operators, improving segmentation accuracy and completeness, and reducing the complexity of segmentation.
Abstract: A CT image-based blood vessel three-dimensional segmentation method includes 1) reading CT contrastographic picture data, marking a layer first with blood vessels as the first layer and generating first layer seed points; 2) growing layer by layer from the second layer The method in the first step includes acquiring the average value of gray values of the first layer; marking the central point of the image of the first layer as the first seed point, growing in the first layer and growing points adjacent to the seed point and whose gray values greater than the average value to be seed points The method in the second step includes calculating the average values of gray values of seed points and non-seed points of the last layer respectively and performing weighing calculation of the average values so as to obtain the minimum gray value Mi of the seed points in the current layer; growing to the current layer from the seed points of the last layer; and growing layer by layer as the above-mentioned way until no new seed points generate any more The collection of all the seed point forms a blood vessel three-dimensional segmentation image The invention aims to provide the CT image-based blood vessel three-dimensional segmentation method reducing dependence on experience of operators, improving segmentation accuracy and completeness

4 citations

Patent
22 Mar 2017
TL;DR: In this paper, a segmentation method and a device for disease speckles on the edge-blurred leaves of protected-cultivation vegetables is presented. And the method comprises the steps of acquiring the gray values of any target point and the gray value of any adjacent point of the target point in a to-be-processed image; according to the grey values of all adjacent points of any point, obtaining an average grey value of the neighborhood of a target point.
Abstract: The present invention provides a segmentation method and a device for disease speckles on the edge-blurred leaves of protected-cultivation vegetables. The method comprises the steps of acquiring the gray value of any target point and the gray value of any adjacent point of the target point in a to-be-processed image; according to the gray values of all adjacent points of any target point, obtaining an average gray value of the neighborhood of the target point; according to k and the transverse and longitudinal coordinate offsets of any adjacent point of any target point, obtaining the weight value of any adjacent point of the target point; according to the gray value of all adjacent points of any target point and the above weight value, obtaining the weighted grayscale mid-value of the neighborhood of the target point; according to the obtained three values of the target point, establishing a three-dimensional histogram (hereinafter referred to as the histogram) and figuring out the region of any target point in the three-dimensional coordinate system (hereinafter referred to as a coordinate system) of the histogram; correcting the three values of the target pint according to a preset strategy; calculating the distance between any target point and the origin of the coordinate system according to the corrected values and conducting the dimension reduction of the histogram according to the distance; segmenting the histogram to obtain an optimal threshold based on OTSU; and segmenting the to-be-processed image according to the threshold. According to the technical scheme of the invention, the computer-aided segmentation for disease speckles is realized.

1 citations