CN110349094A - It is peeled off based on statistics and the 3D point cloud denoising method of adaptive bilateral mixed filtering - Google Patents
It is peeled off based on statistics and the 3D point cloud denoising method of adaptive bilateral mixed filtering Download PDFInfo
- Publication number
- CN110349094A CN110349094A CN201910506642.1A CN201910506642A CN110349094A CN 110349094 A CN110349094 A CN 110349094A CN 201910506642 A CN201910506642 A CN 201910506642A CN 110349094 A CN110349094 A CN 110349094A
- Authority
- CN
- China
- Prior art keywords
- point
- point cloud
- formula
- denoising
- cloud data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000002146 bilateral effect Effects 0.000 title claims abstract description 42
- 238000001914 filtration Methods 0.000 title claims abstract description 39
- 230000003044 adaptive effect Effects 0.000 title claims abstract description 34
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000009499 grossing Methods 0.000 claims abstract description 5
- 239000013598 vector Substances 0.000 claims description 12
- 238000012545 processing Methods 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 7
- 230000000717 retained effect Effects 0.000 claims description 3
- 230000011218 segmentation Effects 0.000 abstract description 3
- 238000000605 extraction Methods 0.000 abstract description 2
- 230000000694 effects Effects 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000000513 principal component analysis Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000003746 surface roughness Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
- G06T2207/20028—Bilateral filtering
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention discloses being peeled off based on statistics and the 3D point cloud denoising method of adaptive bilateral mixed filtering, firstly, calculating in 3D point cloud each point to the average distance of its nearest K neighborhood point, desired value, standard deviation and the distance threshold of average distance；If being filtered out secondly, judgement average distance is greater than distance threshold, otherwise, retain, the 3D point cloud data after being denoised for the first time；Finally, calculating the bilateral smothing filtering factor, and the 3D point cloud data after first denoising, the 3D point cloud data after finally being denoised are handled using the smoothing filter.The method of the present invention is will to count to peel off to eliminate filter and adaptive two-sided filter is effectively combined and carries out denoising to 3D point cloud data.The result shows that the inventive method does not remove only the noise point that peels off, and also remove fluctuating noise point.At the same time, the boundary of 3D point cloud data is smoother after denoising, lays a good foundation for the segmentation and feature extraction of subsequent 3D point cloud data.
Description
Technical field
The invention belongs to 3D point cloud technical field of data processing, and in particular to one kind is peeled off and adaptive bilateral based on statistics
The 3D point cloud denoising method of mixed filtering.
Background technique
With the development of artificial intelligence, machine vision is gradually from two dimensional image excessively to 3-D image, and 3D point cloud conduct
One of Typical Representative of 3-D image is gradually widely used.Under normal conditions, it is obtained using 3D point cloud data acquisition facility
When 3D point cloud, due to the shadow of the factors such as measurand external waviness, surface roughness, equipment precision, ambient lighting, artificial disturbance
It rings, 3D point cloud obtained inevitably by noise influences.The presence of noise can not only seriously affect subsequent 3D point cloud
The pretreatment such as simplification, registration of data, and directly affect precision, 3D point cloud data segmentation effect and the 3D point cloud number of 3D reconstruct
According to the height of discrimination.So whether denoising method selection properly plays decisive role to follow-up work.
By development in recent years, the denoising of 3D point cloud data also emerges numerous research methods.Statistics, which peels off, eliminates filter
Wave algorithm is to calculate point to the distance between distributed point in K neighborhood, and by for statistical analysis to every vertex neighborhood, deletion does not meet mark
Quasi- outlier.Outlier mainly is filtered out, fluctuating noise not can be removed, the boundary after denoising is unsmooth.Bilateral filtering algorithm
It is mainly used for being removed the fluctuating noise of the small scale in 3D point cloud, and can reach the effect of smooth boundary.But the calculation
The removal effect of fado outlier noise is poor.
Summary of the invention
The object of the present invention is to provide a kind of 3D point cloud denoising sides to be peeled off based on statistics with adaptive bilateral mixed filtering
Method solves existing single statistics and peels off that only to filter out the noise point that peels off, boundary unsmooth and single adaptive bilateral for filter
Filter only filters out the problem of fluctuating noise point.
It is denoised the technical scheme adopted by the invention is that being peeled off based on statistics with the 3D point cloud of adaptive bilateral mixed filtering
Method is specifically implemented according to the following steps:
Step 1, it is peeled off using statistics and eliminates filter to the progress denoising of 3D point cloud data, after being denoised for the first time
3D point cloud data；
Step 2, the result in step 1 is denoised using adaptive two-sided filter, the 3D after finally being denoised
Point cloud data.
The features of the present invention also characterized in that
In step 1, it is specifically implemented according to the following steps:
Step 1.1, average distance d of each point to its nearest K neighborhood point in calculating 3D point cloud datam, such as formula (1) institute
Show；
Wherein, K=100, diThe point distance is arrived for every in K neighborhood；
Step 1.2, after step 1.1, the average distance d of each point is calculated separatelymDesired value dpWith standard deviation s；Such as formula
(2) and shown in formula (3)；
Wherein, n is point total quantity, n=958166；
Step 1.3, after step 1.2, the distance threshold d of each point is calculatedtIf dm>dt, then the point is filtered out, otherwise, is protected
It stays；As shown in formula (4)；
dt=dp+λ*s (4)；
In formula (4), λ is relaxation parameter, λ=0.1；dpFor desired value；S is standard deviation.
In step 2, it is specifically implemented according to the following steps:
Step 2.1, the bilateral smoothing factor of each 3D point cloud data after calculating first denoising；As shown in formula (5)；
Wherein, NK(pi) it is piK neighborhood point converge conjunction；pjFor piK neighborhood in a bit；<,>indicate inner product of vectors；nj、
niFor the unit normal vector of corresponding points；WcFor the weighting function of spatial domain；WsFor the weighting function of frequency domain；
Step 2.2, each 3D point cloud data after denoising for the first time in step 1.3 are carried out using formula (9) adaptive
Bilateral filtering processing, the 3D point cloud data after final denoising can be obtained；
pinew=pi+μni(9)；
pinewFor the data point after bilateral filtering, piFor a bit, μ is the bilateral filtering factor, n in cloudiFor piCorresponding list
Position normal vector.
In step 2.1, WcCalculation formula, as shown in formula (6)；
In formula (6), σcFor point piAnd impact factor of the distance to the point between each point in K neighborhood.
In step 2.1, WsCalculation formula, as shown in formula (7)；
In formula (7), σsFor point piImpact factor of the normal direction difference to the point, σ between each point in its K neighborhoodsBy formula
(8) it is calculated；
In formula (8),For piThe average distance put in its K neighborhood, djFor piTo pjDistance.
The beneficial effects of the present invention are:
This method compared with single statistics peels off the effect after eliminating filter and single adaptive two-sided filter denoising,
The noise point of discrete group is not removed only, and eliminates fluctuating noise point.At the same time, 3D point cloud side after denoising is also retained
Boundary it is smooth.The advantages of retaining two kinds of denoising methods, it is existing insufficient to solve them.
Detailed description of the invention
Fig. 1 is the flow chart to be peeled off the present invention is based on statistics with the 3D point cloud denoising method of adaptive bilateral mixed filtering；
Fig. 2 is former 3D point cloud data image to be processed in the present embodiment；
Fig. 3 is to peel off to eliminate filter to the Central Plains Fig. 2 3D point cloud data image denoising figure using statistics in the present embodiment；
Fig. 4 is using adaptive two-sided filter in the present embodiment to the Central Plains Fig. 2 3D point cloud data image denoising figure；
Fig. 5 is peeled off with adaptive bilateral mixed filtering using statistics to the Central Plains Fig. 2 3D point cloud datagram in the present embodiment
As denoising figure.
Specific embodiment
The following describes the present invention in detail with reference to the accompanying drawings and specific embodiments.
The present invention is based on the 3D point cloud denoising method to peel off with adaptive bilateral mixed filtering is counted, as shown in Figure 1, specifically
It follows the steps below to implement:
Step 1, it is peeled off using statistics and eliminates filter to the progress denoising of 3D point cloud data, after being denoised for the first time
3D point cloud data；It is specifically implemented according to the following steps:
Step 1.1, average distance d of each point to its nearest K neighborhood point in calculating 3D point cloud datam, such as formula (1) institute
Show；
Wherein, K=100, diThe point distance is arrived for every in K neighborhood；
Step 1.2, after step 1.1, the average distance d of each point is calculated separatelymDesired value dpWith standard deviation s；Such as formula
(2) and shown in formula (3)；
Wherein, n is point total quantity, and it is bigger to put the more big nearest K neighborhood value being often arranged of total quantity n by n=958166；
Step 1.3, after step 1.2, the distance threshold d of each point is calculatedtIf dm>dt, then the point is filtered out, otherwise, is protected
It stays；As shown in formula (4)；
dt=dp+λ*s (4)；
In formula (4), λ is relaxation parameter, λ=0.1；dpFor desired value；S is standard deviation；
Being peeled off using statistics and eliminating filtering is to guarantee a point cloud feature invariant to the purpose that 3D point cloud data are filtered
Under the premise of can filter out the noise point cloud data that peels off, denoise lay the foundation again for subsequent point cloud data；
Step 2, the result in step 1 is denoised using adaptive two-sided filter, the 3D after finally being denoised
Point cloud data；It is specifically implemented according to the following steps:
Step 2.1, the bilateral smoothing factor of each 3D point cloud data after calculating first denoising；As shown in formula (5)；
Wherein, NK(pi) it is piK neighborhood point converge conjunction, pjFor piK neighborhood in a bit,<,>indicate inner product of vectors, nj、
niFor the unit normal vector of corresponding points, can be calculated by Principal Component Analysis, WcFor the weighting function of spatial domain, bilateral filtering is represented
Smoothness, WsFor the weighting function of frequency domain, the feature reservation degree of bilateral filtering is represented, both passes through gaussian kernel function table
Show；
Wherein, WcCalculation formula, as shown in formula (6)；
In formula (6), σcFor point piAnd impact factor of the distance to the point between each point in K neighborhood；With σcValue increases,
Smoothness is higher, but the reservation degree of feature is reduced；σcBe the distance between two points of lie farthest away in K neighborhood two/
One；
WsCalculation formula, as shown in formula (7)；
In formula (7), σsFor point piNormal direction difference is to the impact factor of the point between each point in its K neighborhood, it and it is bilateral
The feature reserve capability of filtering is directly proportional；σsIt is calculated by formula (8)；
In formula (8),For piThe average distance put in its K neighborhood, djFor piTo pjDistance；
Step 2.2, each 3D point cloud data after denoising for the first time in step 1.3 are carried out using formula (9) adaptive
Bilateral filtering processing, the 3D point cloud data after final denoising can be obtained；
pinew=pi+μni(9)；
pinewFor the data point after bilateral filtering, piFor a bit, μ is the bilateral filtering factor in cloud, formula (5) can be passed through
It calculates, niFor piCorresponding unit normal vector；
Method of the invention is united compared with the sharp group filter of single statistics and single adaptive two-sided filter denoising method
After the 3D point cloud data for counting sharp group filter processing, the 3D point cloud noise that part peels off only is got rid of, not can be removed fluctuating
Noise point, and the boundary after denoising is unsmooth；After the 3D point cloud data of adaptive two-sided filter processing, only get rid of
The 3D point cloud noise of volt not can be removed the 3D point cloud noise for dropping off group.The present invention is will to count peel off filter and adaptive pair
Side filter is effectively combined processing 3D point cloud data, can not only effectively get rid of the 3D point cloud noise to peel off, and
And it can also be effectively removed the 3D point cloud noise of fluctuating, at the same time, treated, and 3D point cloud boundary is more smooth, is subsequent
Good basis is established in the segmentation and feature extraction of 3D point cloud data.
Embodiment
A kind of 3D point cloud denoising method to be peeled off based on statistics with adaptive bilateral mixed filtering, specifically according to the following steps
Implement:
Step 1, it is peeled off using statistics and eliminates filter to the progress denoising of 3D point cloud data, after being denoised for the first time
3D point cloud data；It is specifically implemented according to the following steps:
Step 1.1, the average distance d of each point in formula (1) calculating 3D point cloud data to its nearest K neighborhood point is utilizedm；
Wherein, K is the nearest K neighborhood value of K manually set, K=100, diThe point distance is arrived for every in K neighborhood；
Step 1.2, the average distance d of each point is calculated separately using formula (2) and (3)mDesired value dpWith standard deviation s；
Wherein, n is point total quantity, n=958166；
Step 1.3, distance threshold d is calculated using formula (4)t；If dm>dt, then the point is filtered out, otherwise, is retained；
dt=dp+λ*s (4)；
In formula (4), λ is relaxation parameter, λ=0.1；dpFor desired value；S is standard deviation.
It is peeled off using statistics and eliminates filter denoising is filtered to Fig. 2 original 3D point cloud data, as a result such as Fig. 3 institute
Show, from the figure 3, it may be seen that the tonal noise point around in Fig. 2 is almost filtered out, this lays the foundation for the subsequent noise of removal again；
Step 2, adaptive two-sided filter is recycled to denoise the result in step 1, after finally being denoised
3D point cloud data；It is specifically implemented according to the following steps:
Step 2.1, bilateral smoothing factor is calculated using formula (5)；
Wherein, NK(pi) it is piK neighborhood point converge conjunction, pjFor piK neighborhood in a bit,<,>indicate inner product of vectors, nj、
niFor the unit normal vector of corresponding points, can be calculated by Principal Component Analysis, WcFor the weighting function of spatial domain, bilateral filtering is represented
Smoothness, WsFor the weighting function of frequency domain, the feature reservation degree of bilateral filtering is represented, both passes through gaussian kernel function table
Show；
Wc, calculated using formula (6)；
Wherein, σcFor point piAnd impact factor of the distance to the point between each point in K neighborhood.With σcValue increases, smooth
Degree is higher, but the reservation degree of feature is reduced.σcIt is the half of the distance between two points of lie farthest away in K neighborhood.
Ws, calculated using formula (7)；
Wherein, σsFor point piNormal direction difference is to the impact factor of the point between each point in its K neighborhood, it and bilateral filter
The feature reserve capability of wave is directly proportional；σ is calculated by formula (8)s；
Wherein,For piThe average distance put in its K neighborhood, djFor piTo pjDistance；
Step 2.2, adaptive bilateral filtering processing is carried out using result of the formula (9) to step 1.3.
pinew=pi+μni(9)；
pinewFor the data point after bilateral filtering, piFor a bit, μ is the bilateral filtering factor in cloud, formula (5) can be passed through
It calculates, niFor piCorresponding unit normal vector；
Denoising is filtered to Fig. 2 original 3D point cloud data using adaptive two-sided filter, as a result as shown in figure 4,
As shown in Figure 4, Fig. 2 mesorelief noise point is almost filtered out, but the noise point that peels off cannot remove well.Utilize adaptive pair
Side filter is filtered denoising to the 3D point cloud data of Fig. 3 again, as a result as shown in figure 5, as shown in Figure 5, in Fig. 2
It peels off noise point and fluctuating noise point be effectivelys removed, and the boundary of the 3D point cloud data after denoising is smoother, eliminates
Burr realizes satisfactory visual effect.
Using Y-PSNR, filtering method and the adaptive bilateral filtering method of peeling off are counted by comparing, it can be seen that
The method of the present invention is superior to both the above denoising method.Meanwhile these three methods have handled 50 width different scenes, such as 1 institute of table
Show, the results showed that, the method for the present invention averagely improves 1.0dB than counting the filtering method that peels off, than adaptive bilateral filtering method
Averagely improve 2.3dB.It can be seen that the present invention, which denoises effect, is superior to above two denoising method.
1 different scenes difference denoising method of table denoises Contrast on effect
Claims (5)
1. being peeled off based on statistics and the 3D point cloud denoising method of adaptive bilateral mixed filtering, which is characterized in that specifically according to
Lower step is implemented:
Step 1, peeled off the 3D point for eliminating filter to the progress denoising of 3D point cloud data, after being denoised for the first time using statistics
Cloud data；
Step 2, the result in step 1 is denoised using adaptive two-sided filter, the 3D point cloud after finally being denoised
Data.
2. the 3D point cloud denoising method according to claim 1 to be peeled off based on statistics with adaptive bilateral mixed filtering,
It is characterized in that, in the step 1, is specifically implemented according to the following steps:
Step 1.1, average distance d of each point to its nearest K neighborhood point in calculating 3D point cloud datam, as shown in formula (1)；
Wherein, K=100, diThe point distance is arrived for every in K neighborhood；
Step 1.2, after step 1.1, the average distance d of each point is calculated separatelymDesired value dpWith standard deviation s；Such as formula (2)
And shown in formula (3)；
Wherein, n is point total quantity, n=958166；
Step 1.3, after step 1.2, the distance threshold d of each point is calculatedtIf dm>dt, then the point is filtered out, otherwise, is retained；Such as
Shown in formula (4)；
dt=dp+λ*s (4)；
In formula (4), λ is relaxation parameter, λ=0.1；dpFor desired value；S is standard deviation.
3. the 3D point cloud denoising method according to claim 1 to be peeled off based on statistics with adaptive bilateral mixed filtering,
It is characterized in that, in the step 2, is specifically implemented according to the following steps:
Step 2.1, the bilateral smoothing factor of each 3D point cloud data after calculating first denoising；As shown in formula (5)；
Wherein, NK(pi) it is piK neighborhood point converge conjunction；pjFor piK neighborhood in a bit；<,>indicate inner product of vectors；nj、niIt is right
The unit normal vector that should be put；WcFor the weighting function of spatial domain；WsFor the weighting function of frequency domain；
Step 2.2, each 3D point cloud data after denoising for the first time in step 1.3 are carried out using formula (9) adaptive bilateral
Filtering processing, the 3D point cloud data after final denoising can be obtained；
pinew=pi+μni(9)；
pinewFor the data point after bilateral filtering, piFor a bit, μ is the bilateral filtering factor, n in cloudiFor piCorresponding per unit system
Vector.
4. the 3D point cloud denoising method according to claim 3 to be peeled off based on statistics with adaptive bilateral mixed filtering,
It is characterized in that, in the step 2.1, WcCalculation formula, as shown in formula (6)；
In formula (6), σcFor point piAnd impact factor of the distance to the point between each point in K neighborhood.
5. the 3D point cloud denoising method according to claim 3 to be peeled off based on statistics with adaptive bilateral mixed filtering,
It is characterized in that, in the step 2.1, WsCalculation formula, as shown in formula (7)；
In formula (7), σsFor point piImpact factor of the normal direction difference to the point, σ between each point in its K neighborhoodsBy formula (8)
It is calculated；
In formula (8),For piThe average distance put in its K neighborhood, djFor piTo pjDistance.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910506642.1A CN110349094A (en) | 2019-06-12 | 2019-06-12 | It is peeled off based on statistics and the 3D point cloud denoising method of adaptive bilateral mixed filtering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910506642.1A CN110349094A (en) | 2019-06-12 | 2019-06-12 | It is peeled off based on statistics and the 3D point cloud denoising method of adaptive bilateral mixed filtering |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110349094A true CN110349094A (en) | 2019-10-18 |
Family
ID=68181825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910506642.1A Pending CN110349094A (en) | 2019-06-12 | 2019-06-12 | It is peeled off based on statistics and the 3D point cloud denoising method of adaptive bilateral mixed filtering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110349094A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111722201A (en) * | 2020-06-23 | 2020-09-29 | 常州市贝叶斯智能科技有限公司 | Data denoising method for indoor robot laser radar |
| CN112184556A (en) * | 2020-10-28 | 2021-01-05 | 万申(北京)科技有限公司 | Super-resolution imaging method based on oral CBCT (cone beam computed tomography) reconstruction point cloud |
| CN112396567A (en) * | 2020-11-26 | 2021-02-23 | 河北工业大学 | Scattered point cloud denoising method based on two-step method of normal correction and position filtering |
| CN112907480A (en) * | 2021-03-11 | 2021-06-04 | 北京格灵深瞳信息技术股份有限公司 | Point cloud surface ripple removing method and device, terminal and storage medium |
| CN114359089A (en) * | 2021-12-29 | 2022-04-15 | 江苏苏彩信息技术有限公司 | Three-dimensional point cloud data denoising method based on point cloud filter network |
| CN114897026A (en) * | 2022-05-24 | 2022-08-12 | 上海枢光科技有限公司 | Point cloud filtering method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140267614A1 (en) * | 2013-03-15 | 2014-09-18 | Seiko Epson Corporation | 2D/3D Localization and Pose Estimation of Harness Cables Using A Configurable Structure Representation for Robot Operations |
| CN105096268A (en) * | 2015-07-13 | 2015-11-25 | 西北农林科技大学 | Denoising smoothing method of point cloud |
| CN106599129A (en) * | 2016-12-02 | 2017-04-26 | 山东科技大学 | Multi-beam point cloud data denoising method considering terrain characteristics |
| CN106774942A (en) * | 2017-01-18 | 2017-05-31 | 华南理工大学 | A kind of real-time 3D remote human-machines interactive system |
| CN108986048A (en) * | 2018-07-18 | 2018-12-11 | 大连理工大学 | Based on the quick compound filter processing method of line laser structured light three-dimensional point cloud |
| CN109741265A (en) * | 2018-11-20 | 2019-05-10 | 哈尔滨理工大学 | A kind of spine CT image bilateral filtering denoising method based on feature selecting |
-
2019
- 2019-06-12 CN CN201910506642.1A patent/CN110349094A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140267614A1 (en) * | 2013-03-15 | 2014-09-18 | Seiko Epson Corporation | 2D/3D Localization and Pose Estimation of Harness Cables Using A Configurable Structure Representation for Robot Operations |
| CN105096268A (en) * | 2015-07-13 | 2015-11-25 | 西北农林科技大学 | Denoising smoothing method of point cloud |
| CN106599129A (en) * | 2016-12-02 | 2017-04-26 | 山东科技大学 | Multi-beam point cloud data denoising method considering terrain characteristics |
| CN106774942A (en) * | 2017-01-18 | 2017-05-31 | 华南理工大学 | A kind of real-time 3D remote human-machines interactive system |
| CN108986048A (en) * | 2018-07-18 | 2018-12-11 | 大连理工大学 | Based on the quick compound filter processing method of line laser structured light three-dimensional point cloud |
| CN109741265A (en) * | 2018-11-20 | 2019-05-10 | 哈尔滨理工大学 | A kind of spine CT image bilateral filtering denoising method based on feature selecting |
Non-Patent Citations (3)
| Title |
|---|
| FU YAN: "Research on scattered points cloud denoising algorithm", 《2015 IEEE INTERNATIONAL CONFERENCE ON SIGNAL PROCESSING, COMMUNICATIONS AND COMPUTING (ICSPCC)》 * |
| 李鹏飞: "点云模型的噪声分类去噪算法", 《万方数据知识服务平台》 * |
| 袁华: "基于噪声分类的双边滤波点云去噪算法", 《万方数据知识服务平台》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111722201A (en) * | 2020-06-23 | 2020-09-29 | 常州市贝叶斯智能科技有限公司 | Data denoising method for indoor robot laser radar |
| CN112184556A (en) * | 2020-10-28 | 2021-01-05 | 万申(北京)科技有限公司 | Super-resolution imaging method based on oral CBCT (cone beam computed tomography) reconstruction point cloud |
| CN112184556B (en) * | 2020-10-28 | 2021-10-29 | 万申(北京)科技有限公司 | Super-resolution imaging method based on oral CBCT (cone beam computed tomography) reconstruction point cloud |
| CN112396567A (en) * | 2020-11-26 | 2021-02-23 | 河北工业大学 | Scattered point cloud denoising method based on two-step method of normal correction and position filtering |
| CN112396567B (en) * | 2020-11-26 | 2023-06-06 | 河北工业大学 | Scattered point cloud denoising method based on normal correction and position filtering two-step method |
| CN112907480A (en) * | 2021-03-11 | 2021-06-04 | 北京格灵深瞳信息技术股份有限公司 | Point cloud surface ripple removing method and device, terminal and storage medium |
| CN114359089A (en) * | 2021-12-29 | 2022-04-15 | 江苏苏彩信息技术有限公司 | Three-dimensional point cloud data denoising method based on point cloud filter network |
| CN114359089B (en) * | 2021-12-29 | 2022-09-27 | 江苏苏彩信息技术有限公司 | Three-dimensional point cloud data denoising method based on point cloud filter network |
| CN114897026A (en) * | 2022-05-24 | 2022-08-12 | 上海枢光科技有限公司 | Point cloud filtering method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110349094A (en) | It is peeled off based on statistics and the 3D point cloud denoising method of adaptive bilateral mixed filtering | |
| CN110428433B (en) | Canny edge detection algorithm based on local threshold | |
| CN108416789A (en) | Method for detecting image edge and system | |
| CN109377450B (en) | Edge protection denoising method | |
| Hao et al. | Improved self-adaptive edge detection method based on Canny | |
| CN116777941B (en) | Profile contour detection method and system based on machine vision | |
| CN112819845B (en) | Flexible package substrate contour, line width and line distance defect detection method, medium and equipment | |
| CN103870808A (en) | Finger vein identification method | |
| CN109389560A (en) | A kind of adaptive weighted filter image denoising method, device and image processing equipment | |
| CN104463814A (en) | Image enhancement method based on local texture directionality | |
| Bhargava et al. | An adaptive method for edge preserving denoising | |
| CN106023160B (en) | Blast furnace charge level method for detecting image edge and device | |
| CN110599422A (en) | Weighted average salt and pepper noise reduction algorithm based on edge protection | |
| Thakkar et al. | Edge detection techniques using fuzzy thresholding | |
| CN110647843B (en) | Face image processing method | |
| CN113609998A (en) | Bottom layer spot detection method, electronic device and readable storage medium | |
| CN112465845A (en) | Image fuzzy edge information segmentation method based on computer vision | |
| Zhang et al. | Infrared small target detection using modified order morphology and weighted local entropy | |
| Liu et al. | Infrared ship target image smoothing based on adaptive mean shift | |
| Wong et al. | Adaptive monte carlo retinex method for illumination and reflectance separation and color image enhancement | |
| Wang et al. | Design of wavelet denoising and image enhancement algorithm based on MATLAB | |
| Sun et al. | Study on the preprocessing algorithm of transmission lines video monitoring image | |
| Lixin et al. | Segmentation of breast masses using adaptive region growing | |
| Li et al. | An image dehazing method based on atmospheric veil | |
| CN113487496B (en) | Image denoising method, system and device based on pixel type inference |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191018 |