CN112802071B - Three-dimensional reconstruction effect evaluation method and system - Google Patents
Three-dimensional reconstruction effect evaluation method and system Download PDFInfo
- Publication number
- CN112802071B CN112802071B CN202110088691.5A CN202110088691A CN112802071B CN 112802071 B CN112802071 B CN 112802071B CN 202110088691 A CN202110088691 A CN 202110088691A CN 112802071 B CN112802071 B CN 112802071B
- Authority
- CN
- China
- Prior art keywords
- point cloud
- cloud model
- wing
- dimensional reconstruction
- effect evaluation
- 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.)
- Active
Links
- 238000011156 evaluation Methods 0.000 title claims abstract description 127
- 230000000694 effects Effects 0.000 title claims abstract description 125
- 238000000034 method Methods 0.000 claims abstract description 27
- 241000238631 Hexapoda Species 0.000 claims description 23
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 10
- 238000011158 quantitative evaluation Methods 0.000 abstract description 5
- 238000004422 calculation algorithm Methods 0.000 description 8
- 238000004891 communication Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000003190 augmentative effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
- G06T7/344—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention provides a three-dimensional reconstruction effect evaluation method and a system, comprising the following steps: acquiring a true value point cloud model of a target to be evaluated; carrying out point cloud registration on the true value point cloud model and the reconstructed point cloud model of the target to be evaluated; determining a hausdorff distance between the true point cloud model and the reconstructed point cloud model; and determining a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the Haoskov distance. According to the three-dimensional reconstruction effect evaluation method and system, the Hausdorff distance between each point cloud in the reconstruction point cloud model of the target to be evaluated and the point cloud in the truth point cloud model is calculated, so that the three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model is determined according to the size distribution condition of the Hausdorff distance of each point cloud, the objective quantitative evaluation method for the point cloud reconstruction effect in the aspects of accuracy and completeness of the three-dimensional reconstruction point cloud model is provided, and the objectivity and accuracy of the point cloud reconstruction effect evaluation are effectively improved.
Description
Technical Field
The invention relates to the technical field of machine vision, in particular to a three-dimensional reconstruction effect evaluation method and system.
Background
In the field of machine vision, three-dimensional reconstruction is to reconstruct three-dimensional point cloud according to an image shot by a density camera so as to recover three-dimensional information of a target or a scene, and has wide application in the fields of virtual reality and augmented reality.
The model result generated by three-dimensional reconstruction is mainly the space coordinates of the target or scene surface points, and a quantitative evaluation system for the three-dimensional model of the moth insects is lacking in the process of three-dimensional reconstruction of the insects. At present, the three-dimensional reconstruction effect is generally evaluated subjectively through the reconstruction effect of whole or partial details.
Therefore, how to provide an objective three-dimensional reconstruction effect evaluation method, to objectively and quantitatively evaluate a three-dimensional model, so as to ensure the effective application of the three-dimensional reconstruction technology in the field of insect three-dimensional reconstruction, is a problem to be solved by the function of the person skilled in the art.
Disclosure of Invention
Aiming at the defect of low evaluation reliability caused by subjective evaluation of three-dimensional reconstruction effect in the prior art, the embodiment of the invention provides a three-dimensional reconstruction effect evaluation method and system.
The invention provides a three-dimensional reconstruction effect evaluation method, which comprises the following steps: acquiring a true value point cloud model of a target to be evaluated; carrying out point cloud registration on the true point cloud model and the reconstructed point cloud model of the target to be evaluated; determining a hausdorff distance between the true point cloud model and the reconstructed point cloud model; and determining a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the Haoskov distance.
According to the three-dimensional reconstruction effect evaluation method provided by the invention, the calculation formula for determining the Hausdorff distance between the true value point cloud model and the reconstruction point cloud model is as follows:
H(G,R)＝max[h(G,R),h(R,G)]；
Wherein G is a point cloud set of the true point cloud model, G is any point cloud in the true point cloud set G, R is a point cloud set of the reconstructed point cloud model, R is any point cloud in the reconstructed point cloud set R, H (G, R) is a hausdorff distance from the true point cloud set G to the reconstructed point cloud set R, H (R, G) is a hausdorff distance from the reconstructed point cloud set R to the true point cloud set G, and H (G, R) is a set of hausdorff distances from each point cloud in the true point cloud set G to the reconstructed point cloud set R.
According to the three-dimensional reconstruction effect evaluation method provided by the invention, the three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model is determined according to the Haoskov distance, and the three-dimensional reconstruction effect evaluation method comprises the following steps: counting probability distribution of the Haoskov distance in each preset distance interval; and determining the three-dimensional reconstruction effect evaluation result according to the probability distribution.
According to the three-dimensional reconstruction effect evaluation method provided by the invention, after the three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model is determined, the three-dimensional reconstruction effect evaluation method further comprises the following steps: acquiring first phenotype attribute data of the moth insects from the truth point cloud model; acquiring second phenotype attribute data of the moth insects from the reconstruction point cloud model; obtaining a comparison result of the first phenotype attribute data and the second phenotype attribute data; and carrying out secondary evaluation on the evaluation result of the three-dimensional reconstruction effect according to the comparison result.
According to the three-dimensional reconstruction effect evaluation method provided by the invention, the first phenotype attribute data of the moth insects are obtained from the true value point cloud model, and the method comprises the following steps: acquiring a first wing length, a first wing width, a first wing area and a first wing included angle in the first phenotype attribute data; correspondingly, the obtaining the second phenotype attribute data of the moth insects from the reconstruction point cloud model comprises the following steps: and acquiring a second wing length, a second wing width, a second wing area and a second wing included angle in the second phenotype attribute data.
According to the three-dimensional reconstruction effect evaluation method provided by the invention, the comparison result of the first phenotype attribute data and the second phenotype attribute data is obtained, and the method comprises the following steps: and respectively calculating at least one of root mean square error, relative error and decision coefficient between the first wing length and the second wing length, the first wing width and the second wing width, the first wing area and the second wing area, and the first wing included angle and the second wing included angle as the comparison result.
According to the three-dimensional reconstruction effect evaluation method provided by the invention, when the preset distance interval comprises five intervals of [ 0-1.0 mm ], [ 1.0-2.0 mm ], [ 2.0-3.0 mm ] [ 3.0-4.0 mm ], [ 4.0-5.0 mm ], the three-dimensional reconstruction effect evaluation result is determined according to the probability distribution, and the three-dimensional reconstruction effect evaluation method comprises the following steps: and if the distribution probability of the Haoskov distance in the [ 0-1.0 mm ] interval is greater than a preset probability threshold, determining that the three-dimensional reconstruction effect evaluation result is qualified.
The invention also provides a three-dimensional reconstruction effect evaluation system, which comprises: the point cloud acquisition unit is used for acquiring a true value point cloud model of the target to be evaluated; the point cloud matching unit is used for carrying out point cloud registration on the true value point cloud model and the reconstructed point cloud model of the target to be evaluated; the distance operation unit is used for determining the Hausdorff distance between the true value point cloud model and the reconstruction point cloud model; and the effect evaluation unit is used for determining a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the Haoskov distance.
The invention also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor realizes the steps of the three-dimensional reconstruction effect evaluation method according to any one of the above when executing the program.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the three-dimensional reconstruction effect evaluation method as described in any one of the above.
According to the three-dimensional reconstruction effect evaluation method and system, the Hausdorff distance between each point cloud in the reconstruction point cloud model of the target to be evaluated and the point cloud in the truth point cloud model is calculated, so that the three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model is determined according to the size distribution condition of the Hausdorff distance of each point cloud, the objective quantitative evaluation method for the point cloud reconstruction effect in the aspects of accuracy and completeness of the three-dimensional reconstruction point cloud model is provided, and the objectivity and accuracy of the point cloud reconstruction effect evaluation are effectively improved.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a three-dimensional reconstruction effect evaluation method provided by the invention;
FIG. 2 is a schematic structural diagram of the three-dimensional reconstruction effect evaluation system provided by the invention;
fig. 3 is a schematic structural diagram of an electronic device provided by the present invention.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
It should be noted that in the description of embodiments of the present invention, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element. The orientation or positional relationship indicated by the terms "upper", "lower", etc. are based on the orientation or positional relationship shown in the drawings, are merely for convenience of description and to simplify the description, and are not indicative or implying that the apparatus or elements in question must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as limiting the present invention. Unless specifically stated or limited otherwise, the terms "mounted," "connected," and "coupled" are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communication between two elements. The specific meaning of the above terms in the present invention can be understood by those of ordinary skill in the art according to the specific circumstances.
The three-dimensional reconstruction effect evaluation method and system provided by the embodiment of the invention are described below with reference to fig. 1 to 3.
Fig. 1 is a flow chart of the three-dimensional reconstruction effect evaluation method provided by the invention, as shown in fig. 1, including but not limited to the following steps:
Step S1: acquiring a true value point cloud model of a target to be evaluated;
step S2: carrying out point cloud registration on the true value point cloud model and the reconstructed point cloud model of the target to be evaluated;
step S3: determining a Hausdorff distance between the true value point cloud model and the reconstructed point cloud model;
Step S4: and determining a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the Haoskov distance.
In the subsequent embodiments of the three-dimensional reconstruction effect evaluation method provided by the invention, the reconstruction effect of the reconstruction point cloud model of a certain moth body is evaluated as an example, but the method is not considered as a specific limitation of the protection scope of the invention.
According to the invention, the moth body can be shot by using the structured light three-dimensional scanner, so that the geometric structure and appearance data of the moth body can be obtained according to the shot image, and a true value point cloud model of the moth body can be constructed.
Further, a reconstruction point cloud model of the moth body is obtained, and the reconstruction point cloud model is an evaluation object of the invention, namely, the invention evaluates the reconstruction effect of the reconstruction point cloud model by rescanning the true value point cloud model of the moth body and taking the true value point cloud model as a reference.
The coordinate systems used by the true value point cloud model and the reconstructed point cloud model are inconsistent, and the differences in the size, the spatial position and the gesture are necessarily existed, so that the three-dimensional reconstruction effect evaluation method provided by the invention firstly needs to match and align the true value point cloud model and the reconstructed point cloud model after the true value point cloud model and the reconstructed point cloud model of the moth body are respectively obtained, and therefore, the true value point cloud model and the reconstructed point cloud model can completely overlap an intersection area in space.
Optionally, the point cloud registration method provided by the invention can adopt a manual registration mode or an automatic registration mode based on an algorithm or a registration mode combining manual and algorithm.
The algorithm used may be an iterative closest point algorithm (ITERATIVE CLOSEST POINT, IPC algorithm) or a normal distribution transformation algorithm (Normal Distribution Transform, NDT algorithm), which is not particularly limited in this invention.
The hausdorff distance may be used to measure the distance between proper subsets in space, and may be applied to the distance of an edge matching algorithm. Let X and Y be two proper subsets of the metric space M, then the hausdorff distance H (X, Y) is the smallest number r such that the closed neighborhood of X contains Y, which also contains X.
Because the point cloud model is regarded as a truth value set formed by different point clouds, the three-dimensional reconstruction effect evaluation method provided by the invention utilizes the Hastedorff distance to measure the coincidence degree between the point cloud set corresponding to the truth value point cloud model and the point cloud set corresponding to the reconstructed point cloud model, namely, the Hastedorff distance (the size of one Hastedorff distance corresponding to each point cloud) from each point cloud in all the reconstructed point cloud models to the truth value point cloud model is obtained. Generally, the smaller the value of the Hausdorff distance of a certain point cloud in the reconstructed point cloud model is, the higher the matching degree of the point cloud and the true value point cloud model is, so that the three-dimensional reconstruction effect of the reconstructed point cloud model is evaluated according to the distribution condition of the Hausdorff distances of all the point clouds.
Most desirably, if the hausdorff distance of all the point clouds is 0; if the vast majority of the Haoskov distance size distribution of all the point clouds is about 0 value, the three-dimensional reconstruction effect evaluation result can be determined to be excellent; if the vast majority of the Hausdorff distance size distribution of all the point clouds is in the region with larger deviation from the value 0, the three-dimensional reconstruction effect evaluation result can be determined to be unqualified.
According to the three-dimensional reconstruction effect evaluation method and system, the Hausdorff distance between each point cloud in the reconstruction point cloud model of the target to be evaluated and the point cloud in the truth point cloud model is calculated, so that the three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model is determined according to the size distribution condition of the Hausdorff distance of each point cloud, the objective quantitative evaluation method for the point cloud reconstruction effect in the aspects of accuracy and completeness of the three-dimensional reconstruction point cloud model is provided, and the objectivity and accuracy of the point cloud reconstruction effect evaluation are effectively improved.
Based on the foregoing embodiment, as an optional embodiment, the calculation formula for determining the hausdorff distance between the true point cloud model and the reconstructed point cloud model is as follows:
H(G,R)＝max[h(G,R),h(R,G)]；
Wherein G is a point cloud set of the true point cloud model, G is any point cloud in the true point cloud set G, R is a point cloud set of the reconstructed point cloud model, R is any point cloud in the reconstructed point cloud set R, H (G, R) is a hausdorff distance from the true point cloud set G to the reconstructed point cloud set R, H (R, G) is a hausdorff distance from the reconstructed point cloud set R to the true point cloud set G, and H (G, R) is a set of hausdorff distances from each point cloud in the true point cloud set G to the reconstructed point cloud set R.
Through the calculation formula, the three-dimensional reconstruction effect evaluation method provided by the invention not only calculates the Hausdorff distance between each point cloud in the true value point cloud model and the reconstructed point cloud model; meanwhile, the Haoskov distance between the point cloud in the reconstructed point cloud model and the true value point cloud model is calculated; then taking the maximum value between the two as the Haoskov distance between each point cloud in the true value point cloud set and the reconstructed point cloud model; and finally, combining the Haoskov distances corresponding to all the point clouds in the true value point cloud set to form a Haoskov distance set, and evaluating the three-dimensional reconstruction effect of the reconstructed point cloud model according to the size distribution of all the Haoskov distances in the set.
Based on the foregoing embodiment, as an optional embodiment, the determining, according to the magnitude of the hausdorff distance, a three-dimensional reconstruction effect evaluation result of the reconstructed point cloud model includes: counting probability distribution of the Haoskov distance in each preset distance interval; and determining the three-dimensional reconstruction effect evaluation result according to the probability distribution.
Specifically, as the smaller the Haoskov distance is, the better the three-dimensional reconstruction effect is, the invention provides a method for quantitatively determining the three-dimensional reconstruction effect evaluation result according to the size distribution condition of the Haoskov distance, which comprises the following steps:
First, a plurality of continuous preset distance intervals are divided in advance in a positive number interval (because the magnitude of the hausdorff distance is expressed by a positive number), for example, n preset distance intervals such as [0-a 1]、[a1-a2]、[a2-a3]…[an-1-an ] are divided, and the distance units of the intervals can be selected according to the actual situation of an object to be evaluated, for example, cm or mm is adopted.
Further, statistics is carried out on partial situations of Hausdorff distance between the truth point cloud model and the reconstruction point cloud model determined in the step S3 in the n preset distance intervals, and according to the counted results and preset evaluation criteria, a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model is obtained.
Wherein, the preset evaluation criterion may be that when the probability that the magnitude of the haustorium distance is within the interval of [0-a 1 ] is greater than the preset distribution probability, the evaluation result is an equal reconstruction effect; the probability that the magnitude of the Haoskov distance is within the interval of [0-a 1 ] is smaller than the first preset probability; however, when the probability of being located in the [ a 1-a2 ] interval is larger than the preset distribution probability, the evaluation result is a second-class reconstruction effect …, and according to the mode, until the probability that the magnitude of the Haoskov distance is located in the [ a n-1-an ] interval is larger than the preset distribution probability, the evaluation result is an n-class reconstruction effect. Wherein, the first-class reconstruction effect is better than the second-class reconstruction effect, the second-class reconstruction effect is better than the third-class reconstruction effect, and the reconstruction effect such as … n-1 is better than the n-class reconstruction effect.
As an alternative embodiment, the invention provides a specific reconstruction effect evaluation method, which comprises the following steps: in the case that the preset distance interval includes five intervals of [ 0-1.0 mm ], [ 1.0-2.0 mm ], [ 2.0-3.0 mm ] [ 3.0-4.0 mm ], [ 4.0-5.0 mm ], determining the three-dimensional reconstruction effect evaluation result according to the probability distribution, including: and if the distribution probability of the Haoskov distance in the [ 0-1.0 mm ] interval is greater than a preset probability threshold, determining that the three-dimensional reconstruction effect evaluation result is qualified.
For example, the probability distribution of each interval of 0 to 1.0mm, 1.0 to 2.0mm, 2.0 to 3.0mm, 3.0 to 4.0mm and 4.0 to 5.0mm is calculated by the Hausdorff distance between the true point cloud model and the reconstructed point cloud model, and the reconstruction effect is better when the distribution probability of the interval of 0 to 1.0mm is more than 80 percent.
The three-dimensional reconstruction effect evaluation method provided by the invention can quantitatively evaluate the reconstruction accuracy of the three-dimensional point cloud by measuring the distance between the two models by utilizing the Hastedor distance,
Based on the foregoing embodiment, as an optional embodiment, after determining the three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model, in a case where the target to be evaluated is a moth body, the method further includes: acquiring first phenotype attribute data of the moth insects from the truth point cloud model; acquiring second phenotype attribute data of the moth insects from the reconstruction point cloud model; obtaining a comparison result of the first phenotype attribute data and the second phenotype attribute data; and carrying out secondary evaluation on the evaluation result of the three-dimensional reconstruction effect according to the comparison result.
The three-dimensional reconstruction effect evaluation method for the moth bodies provided by the invention specifically comprises the step of further evaluating the previously acquired evaluation result by combining the phenotype attribute data of the moth bodies contained in the point cloud model on the basis of realizing the three-dimensional reconstruction effect evaluation by utilizing the Hausdorff distance between the truth point cloud model and the reconstruction point cloud model provided in the embodiment.
Wherein the phenotype attribute data mainly comprises wing length, wing width, wing area, wing included angle and the like. The implementation step of the secondary evaluation mainly comprises the steps of obtaining first phenotype attribute data of the moth bodies according to three-dimensional point cloud distribution of the moth bodies in a true-value point cloud model; and obtaining second phenotype attribute data of the moth insects according to the three-dimensional point cloud distribution of the moth insects in the reconstructed point cloud model. Further, by comparing the first phenotype attribute data with the second phenotype attribute data, the evaluation result obtained before can be secondarily evaluated based on the obtained comparison result.
As an alternative embodiment, the first phenotypic attribute data may include, but is not limited to, at least one of a first wing length, a first wing width, a first wing area, and a first wing angle; accordingly, the second phenotypic attribute data also includes at least one of a second wing length, a second wing width, a second wing area, and a second wing angle.
Taking the example of determining the first wing length from all three-dimensional point clouds in the truth point cloud model, how to obtain the first phenotype attribute data according to the truth point cloud model and how to obtain the first phenotype attribute data according to the reconstructed point cloud model is described below:
Firstly, inputting a true value point cloud model into a point cloud analysis model, and identifying wing point clouds of moth insects in the point cloud analysis model in a manual calibration mode; and then, calculating the wing length according to the distribution condition of the wing point cloud, and taking the wing length as the first wing length.
Correspondingly, under the condition that the wing point cloud of the moth insects is calibrated, information such as the width of the first wing, the area of the first wing, the included angle of the first wing and the like can be obtained.
Similarly, the wing point clouds of the moth insects are marked in the reconstruction point cloud model by adopting the same method, and then the second wing length, the second wing width, the second wing area and the second wing included angle are counted.
Optionally, before the true point cloud model and the reconstructed point cloud model are calibrated, pre-processing such as denoising, simplification, registration, hole filling and the like can be performed on the point clouds in the two models.
Based on the foregoing embodiments, as an optional embodiment, the obtaining a comparison result of the first phenotypic attribute data and the second phenotypic attribute data includes:
And respectively calculating at least one of root mean square error, relative error and decision coefficient between the first wing length and the second wing length, the first wing width and the second wing width, the first wing area and the second wing area, and the first wing included angle and the second wing included angle as the comparison result.
The root mean square error is also called standard error, and can be used for measuring the deviation between two comparison values, and taking the calculation of the mean square error of the first wing length and the second wing length as an example, the calculation formula is as follows:
wherein RMSE is root mean square error, n is the comparison frequency of the first fin included angle and the second fin included angle, For the first wing length at the t-th comparison,/>The second wing length at the t-th comparison.
Similarly, the root mean square error between the first wing width and the second wing width, the first wing area and the second wing area, and the first wing included angle and the second wing included angle can be calculated by adopting the formula.
The relative error refers to the value obtained by multiplying the ratio of the absolute error caused by the measurement to the true value measured by 100%, expressed as a percentage. Taking the relative error of the first wing length and the second wing length as an example, the calculation formula is as follows:
wherein RE is the relative error, For the first wing length at the t-th comparison,/>The second wing length at the t-th comparison.
Similarly, by adopting the formula, the absolute errors between the first wing width and the second wing width, the first wing area and the second wing area, and the first wing included angle and the second wing included angle can be calculated respectively.
Further, an average error between the first wing length and the second wing length, the first wing width and the second wing width, the first wing area and the second wing area, and the first wing included angle and the second wing included angle may be determined according to the above relative error, and the average error may be used as an option of the comparison result.
The decision coefficient, also called a determinable coefficient, is a numerical feature representing the relationship between a random variable and a plurality of random variables, and is a statistical index reflecting the degree of reliability of the variation of the dependent variable described by the regression model, and can be defined as the ratio of the variation of the independent variable described by all the independent variables in the model to the total variation of the independent variable.
Taking the determination coefficients of the first wing length and the second wing length as examples, the calculation formula is as follows:
Wherein: p i and O i are the first wing length and the second wing length, respectively, at the ith comparison; p avg and Oavg are the average of the first wing length and the average of the second wing length, respectively; n is the first wing length or the number of comparisons of the first wing length.
According to the three-dimensional reconstruction effect evaluation method provided by the invention, wing length a G in a true value point cloud model G of a moth insect body is extracted, wing length a R of a reconstruction point cloud model R is extracted, root mean square error RMSE a, relative error RE a and a determination coefficient R a 2 of a G and a R are calculated. And then the same calculation method is adopted to respectively calculate root mean square error, relative error and determination coefficient of wing width, wing area and wing included angle, and the calculated values are used as comparison results to carry out secondary evaluation on the three-dimensional reconstruction effect evaluation result determined by the Haosduff distance.
The secondary evaluation may be determining the credibility of the evaluation result of the three-dimensional reconstruction effect.
For example, the method of secondary evaluation may include: under the condition that the determined three-dimensional reconstruction effect evaluation result is qualified according to the hausdorff distance between the true value point cloud model and the reconstruction point cloud model, determining that the root mean square error RMSE a is smaller than a first preset threshold value, the relative error RE a is smaller than a second preset threshold value and the decision coefficient R a 2 is larger than a third preset threshold value, and determining that the secondary evaluation result is: and the evaluation result of the three-dimensional reconstruction effect is that the qualified evaluation is credible.
For another example, if the determined three-dimensional reconstruction effect evaluation result is qualified according to the hausdorff distance between the true value point cloud model and the reconstruction point cloud model, if it is determined that the root mean square error RMSE a is greater than the first preset threshold, or the relative error RE a is less than the second preset threshold, or the decision coefficient R a 2 is greater than the third preset threshold, then the secondary evaluation result is determined as: and the evaluation result of the three-dimensional reconstruction effect is that the qualified evaluation is not credible.
The three-dimensional reconstruction effect evaluation method provided by the invention utilizes the phenotype attribute data of the moth bodies to carry out secondary evaluation on the evaluation result of Haosdorff distance identification, and can effectively provide the evaluation precision.
Fig. 2 is a schematic structural diagram of a three-dimensional reconstruction effect evaluation system provided by the present invention, as shown in fig. 2, including but not limited to a point cloud acquisition unit 21, a point cloud matching unit 22, a distance calculation unit 23, and an effect evaluation unit 24, wherein:
The point cloud acquisition unit 21 is mainly used for acquiring a true value point cloud model of an object to be evaluated; the point cloud matching unit 22 is mainly configured to perform point cloud registration on the true point cloud model and the reconstructed point cloud model of the target to be evaluated; the distance operation unit 23 is mainly used for determining the Hausdorff distance between the true point cloud model and the reconstruction point cloud model; the effect evaluation unit 24 is mainly configured to determine a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the hausdorff distance.
Specifically, in the three-dimensional reconstruction effect evaluation system provided by the invention, the point cloud acquisition unit 21 is used for controlling the structured light three-dimensional scanner to shoot the moth bodies so as to acquire geometric construction and appearance data of the moth bodies, and a true value point cloud model of the moth bodies is constructed.
Further, the photographed true-value point cloud model and the reconstructed point cloud model to be evaluated are matched and aligned by the point cloud matching unit 22, so that the true-value point cloud model and the reconstructed point cloud model complete the complete overlapping of the intersection region in space.
Further, the hausdorff distance from each point cloud in all the reconstructed point cloud models to the true point cloud model is calculated by the distance calculation unit 23, and a hausdorff distance set is constructed.
Finally, the effect evaluation unit 24 is used for analyzing the distribution condition of the sizes of all the hausdorff distances in the hausdorff distance set, and evaluating the three-dimensional reconstruction effect of the reconstruction point cloud model according to the probability of distribution.
According to the three-dimensional reconstruction effect evaluation system provided by the invention, the Hausdorff distance between each point cloud in the reconstruction point cloud model of the target to be evaluated and the point cloud in the truth point cloud model is calculated, so that the three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model is determined according to the size distribution condition of the Hausdorff distance of each point cloud, an objective quantitative evaluation method for the point cloud reconstruction effect in the aspects of accuracy and completeness of the three-dimensional reconstruction point cloud model is provided, and the objectivity and accuracy of point cloud reconstruction effect evaluation are effectively improved.
It should be noted that, when the three-dimensional reconstruction effect evaluation system provided in the embodiment of the present invention is specifically executed, the three-dimensional reconstruction effect evaluation system may be implemented based on the three-dimensional reconstruction effect evaluation method described in any one of the foregoing embodiments, which is not described in detail in this embodiment.
Fig. 3 is a schematic structural diagram of an electronic device provided by the present invention, and as shown in fig. 3, the electronic device may include: processor 310, communication interface (CommunicationsInterface) 320, memory 330 and communication bus 340, wherein processor 310, communication interface 320 and memory 330 communicate with each other via communication bus 340. The processor 310 may invoke logic instructions in the memory 330 to perform a three-dimensional reconstruction effect evaluation method comprising: acquiring a true value point cloud model of a target to be evaluated; carrying out point cloud registration on the true value point cloud model and the reconstructed point cloud model of the target to be evaluated; determining a hausdorff distance between the true point cloud model and the reconstructed point cloud model; and determining a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the Haoskov distance.
Further, the logic instructions in the memory 330 described above may be implemented in the form of software functional units and may be stored in a computer-readable storage medium when sold or used as a stand-alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a Read-only memory (ROM), a random access memory (RAM, randomAccessMemory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform the three-dimensional reconstruction effect evaluation method provided by the above methods, the method comprising: acquiring a true value point cloud model of a target to be evaluated; carrying out point cloud registration on the true value point cloud model and the reconstructed point cloud model of the target to be evaluated; determining a hausdorff distance between the true point cloud model and the reconstructed point cloud model; and determining a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the Haoskov distance.
In still another aspect, the present invention also provides a non-transitory computer-readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the three-dimensional reconstruction effect evaluation method provided by the above embodiments, the method comprising: acquiring a true value point cloud model of a target to be evaluated; carrying out point cloud registration on the true value point cloud model and the reconstructed point cloud model of the target to be evaluated; determining a hausdorff distance between the true point cloud model and the reconstructed point cloud model; and determining a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the Haoskov distance.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (7)
1. The three-dimensional reconstruction effect evaluation method is characterized by comprising the following steps of:
Acquiring a true value point cloud model of a target to be evaluated;
carrying out point cloud registration on the true point cloud model and the reconstructed point cloud model of the target to be evaluated;
determining a hausdorff distance between the true point cloud model and the reconstructed point cloud model;
Determining a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the Haoskov distance;
in the case that the target to be evaluated is a moth-like body, the method further includes:
acquiring first phenotype attribute data of the moth insects from the truth point cloud model;
Acquiring second phenotype attribute data of the moth insects from the reconstruction point cloud model;
Obtaining a comparison result of the first phenotype attribute data and the second phenotype attribute data;
performing secondary evaluation on the three-dimensional reconstruction effect evaluation result according to the comparison result;
The obtaining the first phenotype attribute data of the moth insects from the true point cloud model comprises the following steps:
acquiring a first wing length, a first wing width, a first wing area and a first wing included angle in the first phenotype attribute data;
correspondingly, the obtaining the second phenotype attribute data of the moth insects from the reconstruction point cloud model comprises the following steps:
Acquiring a second wing length, a second wing width, a second wing area and a second wing included angle in the second phenotype attribute data;
The obtaining a comparison result of the first phenotype attribute data and the second phenotype attribute data includes:
And respectively calculating at least one of root mean square error, relative error and decision coefficient between the first wing length and the second wing length, the first wing width and the second wing width, the first wing area and the second wing area, and the first wing included angle and the second wing included angle as the comparison result.
2. The three-dimensional reconstruction effect evaluation method according to claim 1, wherein the calculation formula for determining the hausdorff distance between the true point cloud model and the reconstruction point cloud model is:
H(G,R)＝max[h(G,R),h(R,G)]；
Wherein G is a point cloud set of the true point cloud model, G is any point cloud in the true point cloud set G, R is a point cloud set of the reconstructed point cloud model, R is any point cloud in the reconstructed point cloud set R, H (G, R) is a hausdorff distance from the true point cloud set G to the reconstructed point cloud set R, H (R, G) is a hausdorff distance from the reconstructed point cloud set R to the true point cloud set G, and H (G, R) is a set of hausdorff distances from each point cloud in the true point cloud set G to the reconstructed point cloud set R.
3. The method according to claim 1, wherein determining the three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the magnitude of the hausdorff distance includes:
counting probability distribution of the Haoskov distance in each preset distance interval;
and determining the three-dimensional reconstruction effect evaluation result according to the probability distribution.
4. The three-dimensional reconstruction effect evaluation method according to claim 3, wherein in the case where the preset distance interval includes five intervals of [0 to 1.0mm ], [1.0 to 2.0mm ], [2.0 to 3.0mm ], [3.0 to 4.0mm ], [4.0 to 5.0mm ], the determining the three-dimensional reconstruction effect evaluation result according to the probability distribution includes:
and if the distribution probability of the Haoskov distance in the [ 0-1.0 mm ] interval is greater than a preset probability threshold, determining that the three-dimensional reconstruction effect evaluation result is qualified.
5. A three-dimensional reconstruction effect evaluation system, comprising:
the point cloud acquisition unit is used for acquiring a true value point cloud model of the target to be evaluated;
The point cloud matching unit is used for carrying out point cloud registration on the true value point cloud model and the reconstructed point cloud model of the target to be evaluated;
the distance operation unit is used for determining the Hausdorff distance between the true value point cloud model and the reconstruction point cloud model;
The effect evaluation unit is used for determining a three-dimensional reconstruction effect evaluation result of the reconstruction point cloud model according to the Haoskov distance;
in the case that the target to be evaluated is a moth-like body, the method further includes:
acquiring first phenotype attribute data of the moth insects from the truth point cloud model;
Acquiring second phenotype attribute data of the moth insects from the reconstruction point cloud model;
Obtaining a comparison result of the first phenotype attribute data and the second phenotype attribute data;
performing secondary evaluation on the three-dimensional reconstruction effect evaluation result according to the comparison result;
The obtaining the first phenotype attribute data of the moth insects from the true point cloud model comprises the following steps:
acquiring a first wing length, a first wing width, a first wing area and a first wing included angle in the first phenotype attribute data;
correspondingly, the obtaining the second phenotype attribute data of the moth insects from the reconstruction point cloud model comprises the following steps:
Acquiring a second wing length, a second wing width, a second wing area and a second wing included angle in the second phenotype attribute data;
The obtaining a comparison result of the first phenotype attribute data and the second phenotype attribute data includes:
And respectively calculating at least one of root mean square error, relative error and decision coefficient between the first wing length and the second wing length, the first wing width and the second wing width, the first wing area and the second wing area, and the first wing included angle and the second wing included angle as the comparison result.
6. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the three-dimensional reconstruction effect evaluation method steps of any one of claims 1 to 4 when the computer program is executed by the processor.
7. A non-transitory computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when executed by a processor, implements the three-dimensional reconstruction effect evaluation method steps of any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110088691.5A CN112802071B (en) | 2021-01-22 | 2021-01-22 | Three-dimensional reconstruction effect evaluation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110088691.5A CN112802071B (en) | 2021-01-22 | 2021-01-22 | Three-dimensional reconstruction effect evaluation method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112802071A CN112802071A (en) | 2021-05-14 |
| CN112802071B true CN112802071B (en) | 2024-06-11 |
Family
ID=75811213
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110088691.5A Active CN112802071B (en) | 2021-01-22 | 2021-01-22 | Three-dimensional reconstruction effect evaluation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112802071B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113376660B (en) * | 2021-05-20 | 2022-10-18 | 北京航空航天大学 | Self-adaptive ionospheric model integrity monitoring method |
| CN113408916B (en) * | 2021-06-28 | 2023-12-29 | 河南唐都科技有限公司 | Fire-fighting facility detection and field acceptance assessment system based on intelligent AI and mobile APP |
| CN113470002B (en) * | 2021-07-22 | 2023-11-10 | 中国科学院空天信息创新研究院 | Chromatography SAR three-dimensional point cloud reconstruction quality evaluation method and device |
| CN116188350A (en) * | 2021-11-26 | 2023-05-30 | 中兴通讯股份有限公司 | Point cloud quality evaluation method, equipment and storage medium |
| CN114565648A (en) * | 2022-02-28 | 2022-05-31 | 上海商汤临港智能科技有限公司 | Method, device and equipment for evaluating reconstructed parking space and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017027638A1 (en) * | 2015-08-10 | 2017-02-16 | The Board Of Trustees Of The Leland Stanford Junior University | 3d reconstruction and registration of endoscopic data |
| CN108492329A (en) * | 2018-03-19 | 2018-09-04 | 北京航空航天大学 | A kind of Three-dimensional Gravity is laid foundations cloud precision and integrity degree evaluation method |
| CN109801316A (en) * | 2018-12-19 | 2019-05-24 | 中国农业大学 | A kind of top fruit sprayer three-dimensional point cloud automation method for registering and reconstructing method |
| CN110363858A (en) * | 2019-06-18 | 2019-10-22 | 新拓三维技术(深圳)有限公司 | A kind of three-dimensional facial reconstruction method and system |
| KR20200132065A (en) * | 2019-05-15 | 2020-11-25 | 주식회사 씨에스아이비젼 | System for Measuring Position of Subject |
-
2021
- 2021-01-22 CN CN202110088691.5A patent/CN112802071B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017027638A1 (en) * | 2015-08-10 | 2017-02-16 | The Board Of Trustees Of The Leland Stanford Junior University | 3d reconstruction and registration of endoscopic data |
| CN108492329A (en) * | 2018-03-19 | 2018-09-04 | 北京航空航天大学 | A kind of Three-dimensional Gravity is laid foundations cloud precision and integrity degree evaluation method |
| CN109801316A (en) * | 2018-12-19 | 2019-05-24 | 中国农业大学 | A kind of top fruit sprayer three-dimensional point cloud automation method for registering and reconstructing method |
| KR20200132065A (en) * | 2019-05-15 | 2020-11-25 | 주식회사 씨에스아이비젼 | System for Measuring Position of Subject |
| CN110363858A (en) * | 2019-06-18 | 2019-10-22 | 新拓三维技术(深圳)有限公司 | A kind of three-dimensional facial reconstruction method and system |
Non-Patent Citations (4)
| Title |
|---|
| 一种改进匹配点对选取策略的ElasticFusion室内三维重建算法;王玮琦;游雄;杨剑;李钦;;武汉大学学报(信息科学版)(09);全文 * |
| 便携式RGB-D传感器深度测量与建模精度研究;叶勤;刘行;姚亚会;桂坡坡;;同济大学学报(自然科学版)(06);全文 * |
| 图像匹配点云与激光扫描点云的模型重建对比;王震;;矿山测量(04);全文 * |
| 基于几何结构特征的点云表面重建方法;杨振发;万刚;曹雪峰;李锋;谢理想;;系统仿真学报(11);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112802071A (en) | 2021-05-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112802071B (en) | Three-dimensional reconstruction effect evaluation method and system | |
| US9639769B2 (en) | Liveness detection | |
| CN106023298B (en) | Point cloud Rigid Registration method based on local Poisson curve reestablishing | |
| US9311556B2 (en) | Image analysis for making animal measurements including 3-D image analysis | |
| CN109815770B (en) | Two-dimensional code detection method, device and system | |
| CN110599451A (en) | Medical image focus detection positioning method, device, equipment and storage medium | |
| CN110349117B (en) | Infrared image and visible light image fusion method and device and storage medium | |
| EP3353744B1 (en) | Image analysis for making animal measurements including 3-d image analysis | |
| CN109902641B (en) | Semantic alignment-based face key point detection method, system and device | |
| CN104616022B (en) | A kind of sorting technique of near infrared spectrum | |
| CN112884820A (en) | Method, device and equipment for training initial image registration and neural network | |
| CN113139948A (en) | Organ contour line quality evaluation method, device and system | |
| CN109002758B (en) | Face feature point positioning method, device, equipment and storage medium | |
| CN114549513A (en) | Part identification method, part identification device, quality inspection method, electronic device, and storage medium | |
| CN109145752A (en) | For assessing the method, apparatus, equipment and medium of object detection and track algorithm | |
| CN116664585B (en) | Scalp health condition detection method and related device based on deep learning | |
| CN115690546B (en) | Shrimp length measuring method, device, electronic equipment and storage medium | |
| CN114093521B (en) | Random forest based method and system for estimating blood sugar by reconstructing homogenized samples | |
| CN110934565B (en) | Method and device for measuring pupil diameter and computer readable storage medium | |
| CN112733729A (en) | Model training and regression analysis method, device, storage medium and equipment | |
| CN112784887A (en) | Plant life state evaluation method and device | |
| CN111192234B (en) | Image background clutter measurement method based on internal deduction mechanism | |
| CN117876456A (en) | Non-contact fish weight measurement method and device, electronic equipment and medium | |
| CN114663359A (en) | Method for detecting intramuscular fat content of pig and related product | |
| CN114529720A (en) | Interpretability evaluation method and device of neural network, storage medium and equipment |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |