CN110458088A - A kind of forest scenery resources visual quality evaluation method based on image and principal component - Google Patents
A kind of forest scenery resources visual quality evaluation method based on image and principal component Download PDFInfo
- Publication number
- CN110458088A CN110458088A CN201910726844.7A CN201910726844A CN110458088A CN 110458088 A CN110458088 A CN 110458088A CN 201910726844 A CN201910726844 A CN 201910726844A CN 110458088 A CN110458088 A CN 110458088A
- Authority
- CN
- China
- Prior art keywords
- forest
- visual quality
- landscape
- resources
- scenery
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides a kind of forest scenery resources visual quality evaluation method based on image and principal component, using the forest scenery resources in suburban forest park as research object, visual quality evaluation is carried out to forest scenery resources using the image data that unmanned plane and portable hand-held video camera are recorded, the evaluation of estimate of the public is standardized.The sight factor of public's influence forest scenery resources of interest is obtained using questionnaire survey mode；With Principal Component Analysis, dimensionality reduction is carried out to the factor for influencing visual quality using SPSS23.0 software, it is core component which, which is distinguished, finally obtain the visual signature factor for meeting the forest landscape that the public watches demand, respectively forest landscape Nature closeness, forest landscape regional culture feature, forest colorfulness, arbor feature easy to identify, spatial impression in woods, the present invention helps to improve the visual quality of forest scenery resources, provides scientific reference for management, the artificial construction of forest landscape.
Description
Technical field
The invention belongs to forest scenery resources visual evaluation technical fields, and in particular to a kind of based on image and principal component
Forest scenery resources visual quality evaluation method.
Background technique
Landscape resources are also known as landscape resource, Scenic resource, scenery tourist resources, are to refer to that people is caused to examine
Beauty and visit activity, can be used as the general name of the natural resources of development and utilization.Forest scenery resources are wherein to play landscape function
It can be core, the artificial or natural forest vegetation with ecologic stability.It is the fundamental for constituting landscape resources,
It is the material base of scenic spot generation environment benefit, Social benefit and economic benefit.
In recent years along with the development of outdoor travel industry, people to the craving of back to nature, to fine habitat to
Past, forest landscape becomes the hot spot of public attention as the landscape resources with higher ecology, aesthetics and cultureal value.With
Forest landscape investigates work to forest scenery resources more and more by the concern of the public, construction ecologic structure is perfect,
Unique Aesthetic Art, the forest landscape rich in region cultural traits are a core contents of current research.According to 2017 years
National Forest Park operating statistics data, up at 3505, tourist 9.62 is received in national Forest Park altogether for China Forest Park
Hundred million person-times, 878.5 hundred million yuan of the direct tourist earnings drives nearly 880,000,000,000 yuan of social synthesis's income.In June, 2017, the State Administration of Forestry
It prints and distributes " about the instruction for accelerating that suburban forest park is promoted to develop ", the important meaning of further emphasized development Forest Park
Justice.
In general, the domestic positive research about forest scenery resources is less, and the quantitative study of view-based access control model evaluation
Just seem less.Forest scenery resources are tourisms of strolling about or have a rest as natural resources and landscape resources important in suburban forest park
The material base of industry sustainable development plays important ecology, economy and society value, thus the visual quality of its own is commented
Valence is just particularly important.
In conclusion problem of the existing technology is:
In the research of previous landscape evaluation, people are mostly by the way of this two dimensional image of photo as experiment sample
This, although two dimensional image can objective reaction actual conditions to a certain extent, however as science and technology and vision research
Development, has shown that two dimensional image can not accurately reflect the visual perception of actual scene entirety, thus the skill of visual evaluation
Art has progress to be updated.For influence visual evaluation factorial analysis, conventional method mostly use analytic hierarchy process (AHP), Delphi method or
It is Field Using Fuzzy Comprehensive Assessment etc., however the expert of these methods is subjective, exists and is difficult to catch principal contradiction to reflect things
It is not objective enough to will lead to analysis conclusion to a certain extent for the problem of rule between built-in variable.Domestic related forest landscape
Research gradually rose in recent years, in general, positive research at this stage is less, view-based access control model evaluation quantitative study just show
It obtains less.Forest scenery resources as natural resources and landscape resources important in suburban forest park, be stroll about or have a rest tourist industry can
The material base of sustainable development, therefore visual quality evaluation is just particularly important.
Solve the difficulty and meaning of above-mentioned technical problem:
The visual quality evaluation of landscape technically has progress to be updated with the development of science and technology and vision research,
And the visual evaluation for being directed to forest landscape this specific landscape resources is still inadequate, thus needs to be studied.Landscape resources
Visual evaluation be estimator's psychological activity and landscape the interactive result of visual quality.The visual quality of forest landscape is not
It is only influenced by characteristic of plant communities, color aesthetic features, while also by the cultural traits that it is formed in region for a long time
The influence of visual psychology perception factor when being watched with people.Therefore, how personal vision is watched to the operation with forest landscape
Management combines, and serves the promotion of landscape quality, provides data supporting for landscape planning, is it is necessary to the problem of thinking deeply.
The purpose of forest landscape visual quality evaluation study can be summarized as follows: 1. around Forest Park main travel route and sight spot
Visual quality level is improved, is taken a firm foundation for the tourist industry of Forest Park；2. carrying out in forest landscape to adaptation to local conditions
Diversified economy management with local characteristic；3. building the forest landscape sight spot with feature of local culture, meet the local public
Growing amusement and recreation demand plays the function in outskirts of a town forest greenery patches.
In view of the problems of the existing technology, the present invention provides a kind of forest wind based on unmanned plane image and principal component
Scape resource visual quality evaluation method.
Summary of the invention
The forest scenery resources visual quality evaluation method based on image and principal component that the object of the present invention is to provide a kind of,
It is led in a manner of solving traditional this two dimensional image using photo as the factorial analysis of experiment sample and influence visual evaluation
Final forest scenery resources visual quality is caused to evaluate not objective enough and inaccurate problem.
The present invention provides the following technical solutions:
A kind of forest scenery resources visual quality evaluation method based on image and principal component, comprising the following steps:
S1, forest landscape money is collected in such a way that two kinds of equipment of unmanned plane and field camera record 4K image data
The image data in source；S2, visual quality evaluation personnel regard the image data of forest scenery resources using Likert scale
Feel quality evaluation；S3, visual quality evaluation personnel visual quality impact factor of interest is obtained using questionnaire survey mode；
S4, using SPSS23.0 software, dimensionality reduction is carried out to impact factor with Principal Component Analysis, to obtain meeting public's demand
The visual signature factor of forest scenery resources.
Further, in the S1, the unmanned plane is DJI-Mavic 2Pro unmanned plane, and the field camera is
The forest scenery resources that DJI-Osmo pocket field camera, the unmanned plane and the field camera are recorded
Image data is the experiment sample of 4K format；What the forest landscape of the forest scenery resources selected is area for 400 ㎡ (20m*
Forest vegetation 20m), the recording of image data are divided into two sections, respectively include the simulated altitude angle of unmanned plane recording
The scene for simulation people's angle viewing forest landscape that the scene and video camera for watching forest landscape are recorded.
Further, in the S2, the visual quality evaluation personnel for no colour blindness, without the public of anomalous trichromatism, be based on
Social sighting distance expansion forest scenery resources visual quality evaluation, social sighting distance, that is, public 70-100m distance outside woods
It is ornamental.
Further, there was only the room of audio-visual playback equipment between the visual quality evaluation personnel individually will will be brought into one
Interior, evaluation procedure includes following three step:
A, to visual quality evaluation personnel, this time the whole of evaluation content is introduced, and plays all image data,
So that visual quality evaluation personnel can quickly form the overall impression of forest landscape；
B, two sections of image datas for individually playing forest landscape, the visual quality watched for forest landscape make evaluation
Value；Visual quality evaluation personnel uses Likert scale when evaluating to forest landscape, uses " 5,3,1,0, -1, -3, -5 " point
" very high, high, higher, general, lower, low, very low " of visual quality is not corresponded to.
C, evaluation is made for visual quality impact factor, opinion scale positions 1,2,3,4,5 Pyatyi score values.
Further, it is handled for the evaluation that the visual quality is made using Spass23.0 software standardization, the mark
Standardization handles formula are as follows:
Valueij=(Rij-Rj)/Sj；Valuei=∑ Zij/Nj；
Symbol meaning is as follows: ValueijIt is jth position visual quality evaluation personnel to the evaluation criterion of forest landscape at i-th
Value, RijIt is jth position estimator to the evaluation of estimate of forest landscape at i-th；RjIt is jth position visual quality evaluation personnel to all forests
The evaluation average value of landscape, RjIt is poor for evaluation criterion of the jth position estimator to all forest landscapes；ValueiFor forest wind at i-th
The evaluation criterion value of scape, NjFor all estimator's numbers.
Further, for the score value of the visual quality impact factor using the immediate score value of mean value as the vision matter
Measure the scoring criteria of impact factor, formula are as follows:
Fik=∑ Iijk/Njk
Symbol meaning is as follows: FikFor the average value of k visual quality impact factors of i-th piece of forest landscape, IijkFor jth position
Evaluation of estimate of the visual quality evaluation personnel to k visual quality impact factors of i-th piece of forest landscape, NjkFor k metrics evaluations
All estimator's numbers.
Further, the visual quality impact factor include: ecological factor, aesthetic values, cultural words, psychology at
Five cause, time history dimensions.
Further, in the S3, the questionnaire survey mode includes to age bracket, profession, educational background, local and annual
Removing Forest Park number is the visual quality evaluation personnel of background.
Further, in the S3, visual quality evaluation personnel sample is analyzed using the normalized form that Scheaffer is proposed
Whether amount has enough representativenesses:
Wherein n is visual quality evaluation personnel quantity, and N is the questionnaire survey size of population that statistics represents, and δ is that sample misses
Difference, when δ takes 0.05, confidence coefficient >=95%, the Population that the statistics of n >=400 represents at this time；
Utilize the confidence level of the side reaction coefficient verifying questionnaire proposed by Cronbach, calculation formula are as follows:
K indicates the forest landscape number of pictures of questionnaire, SiIndicate the variance of respondent, Sx2What is indicated is all tune
Check the variance of elephant；
Further, in the S4, the visual signature factor of the forest scenery resources includes forest landscape Nature closeness
Feature, forest landscape regional culture feature, forest colorfulness feature, arbor feature easy to identify and Lin Nei spatial impression feature；
The factor of the forest landscape Nature closeness feature includes: composition TCS, biotype composition LC, the vertical hierachy number SS of group
With hierarchy correlation degree SC；The factor of the forest landscape regional culture feature includes: indigenous plant amount NPP, folk custom culture connotation
Spend CFC, cooking culture intension degree FCC, religious belief cultural connotation degree CRB, Farming Culture intension degree CFC2 and time history
TH；The factor of the forest colorfulness feature includes aspect tree species ratio PST, pattern richness VDC, leaf color richness
CR, color contrast CC and plant color color CI；The factor of the arbor feature easy to identify includes trunk display degree VS, tree crown
Resolution ITC and group kink characteristics CS；The factor of spatial impression feature includes canopy density CD and intervisibility degree DV in the woods.
The beneficial effects of the present invention are:
A kind of forest scenery resources visual quality evaluation method based on image and principal component of the present invention, with prior art phase
Than the collection method of forest scenery resources data will more objectively reflect the visual quality situation of forest landscape；
Obtain public's forest landscape visual quality evaluation points system of interest based on questionnaire survey, include: it is ecological because
Five son, aesthetic values, cultural words, Mental Origin, time history dimensions, totally 32 factors.With Principal Component Analysis pair
Impact factor carries out dimensionality reduction, and the obtained forest landscape visual signature factor for meeting public's demand is respectively as follows: forest landscape closely certainly
Right degree, forest landscape regional culture feature, forest colorfulness, arbor feature easy to identify, spatial impression in woods.This will be served
The visual quality of forest scenery resources is promoted, and provides data supporting for the landscape planning and planning and designing of forest scenery resources.
Detailed description of the invention
Attached drawing is used to provide further understanding of the present invention, and constitutes part of specification, with reality of the invention
It applies example to be used to explain the present invention together, not be construed as limiting the invention.In the accompanying drawings:
Fig. 1 is appraisal procedure flow diagram of the present invention；
Fig. 2 is the essential information figure of Different Forest landscape sampling point；
Fig. 3 is the essential information figure of visual quality evaluation personnel；
Fig. 4 is visual quality impact factor system figure；
Fig. 5 is the scatter plot that the public for the different majors background chosen scores to forest landscape visual quality；
Fig. 6 is that population variance explains tabular drawing；
Fig. 7 is postrotational component matrix tabular drawing.
Specific embodiment
As shown in Figure 1, a kind of forest scenery resources visual quality evaluation method based on image and principal component, including it is following
Step:
S1, forest landscape money is collected in such a way that two kinds of equipment of unmanned plane and field camera record 4K image data
The image data in source；
Unmanned plane is DJI-Mavic 2Pro unmanned plane, and field camera is DJI-Osmo pocket Portable image pickup
The image data for the forest scenery resources that machine, unmanned plane and field camera are recorded is the experiment sample of 4K format；Forest wind
What the forest landscape of scape resource selected is forest vegetation of the area for 400 ㎡ (20m*20m), the recording point of image data
It is two sections, the simulation people that the scene and video camera for respectively including the simulated altitude angle viewing forest landscape of unmanned plane recording are recorded
The scene of angle viewing forest landscape.
S2, visual quality evaluation personnel carry out visual quality using image data of the Likert scale to forest scenery resources
Evaluation；
Visual quality evaluation personnel is no colour blindness, without the public of anomalous trichromatism, and forest landscape is unfolded based on social sighting distance
The evaluation of resource visual quality, social sighting distance, that is, public outside woods 70-100m distance it is ornamental.Visual quality evaluation personnel will
There was only the interior of audio-visual playback equipment between individually being brought into one, evaluation procedure includes following three step:
A, to visual quality evaluation personnel, this time the whole of evaluation content is introduced, and plays all image data,
So that visual quality evaluation personnel can quickly form the overall impression of forest landscape；
B, two sections of image datas for individually playing forest landscape, the visual quality watched for forest landscape make evaluation
Value；Visual quality evaluation personnel uses Likert scale when evaluating to forest landscape, uses " 5,3,1,0, -1, -3, -5 " point
" very high, high, higher, general, lower, low, very low " of visual quality is not corresponded to.
C, evaluation is made for visual quality impact factor, opinion scale positions 1,2,3,4,5 Pyatyi score values.
It is handled for the evaluation that visual quality is made using Spass23.0 software standardization, standardization formula are as follows:
Valueij=(Rij-Rj)/Sj；Valuei=∑ Zij/Nj；
Symbol meaning is as follows: ValueijIt is jth position visual quality evaluation personnel to the evaluation criterion of forest landscape at i-th
Value, RijIt is jth position estimator to the evaluation of estimate of forest landscape at i-th；RjIt is jth position visual quality evaluation personnel to all forests
The evaluation average value of landscape, RjIt is poor for evaluation criterion of the jth position estimator to all forest landscapes；ValueiFor forest wind at i-th
The evaluation criterion value of scape, NjFor all estimator's numbers.
For visual quality impact factor score value using the immediate score value of mean value as the visual quality impact factor
Scoring criteria, formula are as follows:
Fik=∑ Iijk/Njk
Symbol meaning is as follows: FikFor the average value of k visual quality impact factors of i-th piece of forest landscape, IijkFor jth position
Evaluation of estimate of the visual quality evaluation personnel to k visual quality impact factors of i-th piece of forest landscape, NjkFor k metrics evaluations
All estimator's numbers.
S3, visual quality evaluation personnel visual quality impact factor of interest is obtained using questionnaire survey mode；
Visual quality impact factor includes: ecological factor, aesthetic values, cultural words, Mental Origin, time history five
Dimension；Questionnaire survey mode includes to age bracket, profession, educational background, local and the annual view for removing Forest Park number as background
Feel quality evaluation personnel；
Whether the normalized form analysis visual quality evaluation personnel sample size proposed using Scheaffer has enough generations
Table:
Wherein n is visual quality evaluation personnel quantity, and N is the questionnaire survey size of population that statistics represents, and δ is that sample misses
Difference, when δ takes 0.05, confidence coefficient >=95%, the Population that the statistics of n >=400 represents at this time；
Utilize the confidence level of the side reaction coefficient verifying questionnaire proposed by Cronbach, calculation formula are as follows:
K indicates the forest landscape number of pictures of questionnaire, SiIndicate the variance of respondent, Sx2What is indicated is all tune
Check the variance of elephant；
S4, using SPSS23.0 software, dimensionality reduction is carried out to impact factor with Principal Component Analysis, to obtain meeting public affairs
The visual signature factor of the forest scenery resources of many demands.
The visual signature factor of forest scenery resources includes forest landscape Nature closeness feature, forest landscape regional culture spy
Sign, forest colorfulness feature, arbor feature easy to identify and Lin Nei spatial impression feature；Forest landscape Nature closeness feature because
Son includes: composition TCS, biotype composition LC, group vertical hierachy number SS and hierarchy correlation degree SC；Forest landscape region text
The factor for changing feature includes: indigenous plant amount NPP, folk custom culture connotation degree CFC, cooking culture intension degree FCC, religious belief text
Change intension degree CRB, Farming Culture intension degree CFC2 and time history TH；The factor of forest colorfulness feature includes aspect
Tree species ratio PST, pattern richness VDC, leaf color richness CR, color contrast CC and plant color color CI；Arbor is easily known
The factor of other feature includes trunk display degree VS, crown identification degree ITC and group kink characteristics CS；The factor of spatial impression feature in woods
Including canopy density CD and intervisibility degree DV.
Present embodiment is by taking the Purple Mountain of Nanjing as an example, and steps are as follows,
1), using the method on field work and typical sample ground to expansion investigation early period of Purple Mountain forest scenery resources, selection
At 6 near tourist attractions 48 at sampling of the sampling point as forest landscape, measured and recorded the basic letter on every place's sample ground
Breath；The longitude, latitude and height above sea level on sample ground are had recorded using GPS instrument (Spectra Precision SP60)；Use hand-held
Laser range finder (Bosch Glm150) measures Terrain Elevation and observed range；The slope aspect of landform has been determined with compass；With number
Word photometer (LX-102) measures the illumination of corresponding period.With having recorded investigation same day weather condition and forest landscape sample simultaneously
Plant composition, the tall altitude information (as shown in Figure 2) for filling grass layer.
2), data collection uses DJI unmanned plane (Mavic 2Pro) and DJI field camera (Osmo pocket) two
The forest landscape 4K image that kind of equipment is recorded is used as experiment sample, be based on " social sighting distance " (i.e. the public outside woods 70-100m away from
From it is ornamental) the visual quality evaluations of expansion forest scenery resources.What forest landscape selected is area for 400 ㎡ (20*20m)
Forest vegetation, the recording of image are divided into two sections: first is that unmanned plane was recorded, simulation is high-altitude angle viewing forest landscape
Scene；Second is that video camera record, simulation be people's angle viewing forest landscape scene.
The acquisition of image data follows following principle: 1. forest scenery resources are the plant based on arbor, shrub and vegetation
Group eliminates other non-plant elements such as people, animal and structures；2. image recording condition is based on stringent specific weather
Between condition: recorded between 9:00am to 4:00pm, record day it is fine, 1-2 grades of wind speed, PM2.5 is excellent；3. image recording
Using honourable mode, the resolution ratio of recording is 4K, is shot under the conditions of frontlighting；4. recorded using unmanned plane, tilted using 45 °
Camera work is recorded clockwise around object, recording time 60S；When adopting video camera recorded video, hand-held instrument people view
Height is recorded clockwise around object, and recording time 60S, two ways is controlled in 80m apart from scenic forest.
3) questionnaire, questionnaire design and data source: is carried out to 448 public without visual problems such as colour blindness or anomalous trichromatisms
Investigation, the public is from different age group, different majors, different academic backgrounds, different local and every year removes Forest Park not homogeneous
Several background, to obtain comprehensive initial data source, interviewed person essential information (as shown in Figure 3).
4), the evaluation of landscape resources visual quality is the knot of landscape unique characteristics and the effect of estimator's visual psychology action interactions
Fruit.Visual evaluation is chosen because the period of the day from 11 p.m. to 1 a.m not only allows for forest landscape phytobiocoenose multifactor otherness, while also being analyzed
The visual psychology when cultural traits and people that forest landscape is formed in region for a long time watch perceives factor.It is domestic and international using for reference
On the basis of the research achievements such as the ornamental, Plants Culture of related phytobiocoenose ecological functions structure, plant color, by the public's
Questionnaire, construct comprising ecological factor, aesthetic values, cultural words, Mental Origin, five dimensions of time history evaluation points
System, totally 32 impact factors (as shown in Figure 4).
5) there was only the interior of audio-visual playback equipment, evaluation procedure between, visual quality evaluation personnel individually will will be brought into one
Including following three step: (as shown in Figure 5)
A, to visual quality evaluation personnel, this time the whole of evaluation content is introduced, and plays all image data,
So that visual quality evaluation personnel can quickly form the overall impression of forest landscape；
B, two sections of image datas for individually playing forest landscape, the visual quality watched for forest landscape make evaluation
Value；Visual quality evaluation personnel uses Likert scale when evaluating to forest landscape, uses " 5,3,1,0, -1, -3, -5 " point
" very high, high, higher, general, lower, low, very low " of visual quality is not corresponded to.
C, evaluation is made for visual quality impact factor, opinion scale positions 1,2,3,4,5 Pyatyi score values.
6), visual quality evaluation personnel amount is analyzed: according to the sample availability research of Scheaffer et al., when statistics generation
When the number of table is more than or equal to 400, confidence coefficient is more than or equal to 95% at this time, and the public of this investigation amounts to 448 people, shows to comment
Valence person's sample size is enough.
7), credibility of sample's is analyzed: the confidence level of questionnaire is verified according to a coefficient of Cronbach et al. proposition,
Calculating this evaluation questionnaire a is 0.926 (a > 0.8), shows that evaluation sample interior consistency is good, tests with a high credibility.Sample
Feature description.
It is 8), whether statistically significant using the difference between multivariable duplicate measurements variance analysis visual score standard value,
Inspection result is consistent, and Value value P < 0.001 shows with statistical significance.The analysis of estimator's individual difference the result shows that:
There is no the evaluations to forest landscape visual quality to generate for combination of interactions between the single factor test of public individual and these single factor tests
It significantly affects.
9), factorial analysis；The extraction and name of the factor.KMO and bartlett's test are carried out to the evaluation data of the factor, shown
Show that KMO is 0.732 > 0.6, conspicuousness P is 0.013 < 0.05, can carry out principal component analysis.It explains and is known (such as from population variance
Shown in Fig. 6), the accumulation contribution rate of preceding 5 characteristic values is 86.203%, more than 85%.Extracting method: principal component analysis.This 5
Ingredient can preferably express the entire change rule of original 32 factors, calculate postrotational factor loading as a result, accordingly may be used
Obtain the expression formula of each principal component.
10), spinning solution (as shown in Figure 7): Kaiser standardizes varimax.It has been restrained after being rotated in 6 iteration.
1st ingredient includes 4 factors, and factor load is at 0.630 point or more, respectively in the vertical hierachy number of group
(SS), composition (TCS), biotype constitute that (LC), load is larger on hierarchy correlation degree (SC), wherein the vertical hierachy number of group
(SS) load highest is 0.883 point.These impact factors embody forest landscape close to natural degree, therefore define F1
For forest landscape Nature closeness.
2nd ingredient includes 6 factors, and factor load is at 0.766 point or more, respectively in folk custom culture connotation degree
(CFC), time history (TH), religious belief cultural connotation degree (CRB), indigenous plant amount (NPP), cooking culture intension degree
(FCC), load is larger on Farming Culture intension degree (CFC2), wherein folk custom culture connotation degree (CFC) load highest, is
0.951 point.This embodied a concentrated reflection of that forest landscape constitutes in the mental mechanism of people with the culture that can be recognized, rely on and survive
Speciality, the degree with regional culture intension, therefore defining F2 is forest landscape regional culture feature.
3rd ingredient includes 5 factors, and factor load is at 0.612 point or more, respectively in leaf color richness (CR), flower
Color richness (VDC), aspect tree species ratio (PST), color contrast (CC), load is larger on plant color color (CI), leaf
Color richness (CR) load highest is 0.825 point.Forest colorfulness situation is reflected in these factor sets, therefore fixed
Adopted F3 is forest landscape colorfulness.
4th ingredient includes 3 factors, and factor load is at 0.668 point or more, respectively in crown identification degree (ITC), tree
Load is larger on dry display degree (VS) and arbor kink characteristics (CS), wherein crown identification degree (ITC) load highest, is 0.778
Point.These factor key reactions are whether forest tree species are easy to be known by the public another characteristic, therefore defining F4 is that arbor is easily known
Other feature.
5th ingredient includes 2 factors, and factor load is at 0.731 point or more, respectively in intervisibility degree (DV), canopy density
(CD) load is larger on, wherein intervisibility degree (DV) load highest, is 0.774 point.The forest wind of these factor key reactions
The space sense of vision in scape, therefore defining F5 is spatial impression in woods.
11) following comprehensive characteristics, are established because of subformula according to postrotational component matrix.
F1 (forest landscape Nature closeness)=0.293SS+0.268TCS+0.261LC+0.211SC；
F2 (forest landscape regional culture feature)=0.185CFC+0.182TH+0.171CRB+0.162NPP+0.151FCC
+0.149CFC2；
F3 (forest landscape colorfulness)=0.221CR+0.215VDC+0.210PST+0.190CC+0.164CI；
F4 (arbor feature easy to identify)=0.361ITC+0.329VS+0.310CS；
F5 (spatial impression in woods)=0.514DV+0.486CD.
Forest scenery resources principal component analysis based on public's visual quality.
According to characterization factor formula F 1, in this characterization factor of forest landscape Nature closeness, the vertical level of group is abundant,
The compositions of more tree species, the biotype of tall shrub and scenic forest visual quality with obvious hierarchy correlation are higher, more can be by public affairs
Many likes.This is because people generally more receive have wilderness and Natural Mood in this kind of large natural environment in Forest Park
Forest landscape landscape.Analyze the forest landscape of Purple Mountain visual quality high score, it has been found that forest community can be divided into Qiao
Timber layer, shrub layer and herbaceous layer, and arborous layer can be divided into 3 sub-layers.
1st sub-layer in 15m or more, comprising Chinese sweet gum Liquidambarformosana, Chinese tallow tree Sapium sebiferum,
Celtis julianae Celtis julianae, aphananthe aspera Aphananthe aspera, glossy privet Ligustrum lucidum, quercus chenii
Quercus chenii etc., tree-like tall and straight leaf color is beautiful, configures as forest landscape extended background.
The 2nd high 5-15m of sub-layer, mainly comprising hackberry Celtis sinensis, Bischofia javanica Bl Bischofia polycarpa,
Hair Lai Swidawalteri, Folium Alniphylli fortumei Alniphyllum fortunei, Chinese ash Fraxinus chinensis, kalopanax septemlobus
Kalopanax septemlobus, live oak Quercus dentata etc., tree performance is graceful, leaf color is beautiful, their general scatterplots are matched
It plants, arrange in pairs or groups naturally.
3rd sub-layer 3-5m includes south candle Vaccinium bracteatum, winged euonymus Euonymus alatus, white Du
The plant shapes such as Euonymus maackii, thorny elaeagnus Elaeagnus pungens are compared with dungarunga.Shrub layer is generally in 2m hereinafter, logical
Often configuration large area form protrusion or coloury shrub, including prunus cerasifera Prunus cerasifera, golden-rimmed Chinese littleleaf box
Euonymus japonicu, Loropetalum chinense var.rubrum Loropetalumchinense, Ligustrum quihoui Ligustrum quihoui, narrow leaf
Chinese littleleaf box Buxus stenophylla, Yunnan jasmine Jasminumyunnanense etc..And in arborous layer, shrub layer and herbaceous layer
Outside, forest community interlayer is usually also with liana, such as: trachelospermum jasminoide Trachelospermum jasminoides, Boston ivy
Parthenocissus tricuspidata, climbing fig Ficus pumila, reach the clouds Campsis grandiflora etc..
According to characterization factor formula F 2, influence of the regional culture feature of forest landscape to visual quality occupies second
It sets, the plant in forest landscape produces huge shadow to the thinking activities of people and Cognitive Mode in territorial environment for a long time
It rings, and forms the component part of human culture, this is also just understood that regional culture feature is evaluated in visual quality in people
Occupy the second critical positions.In long history evolution process, Purple Mountain forest landscape forms its distinctive folk culture, ancestor
Teach the features such as belief culture, cooking culture, Farming Culture.The forest landscape on the Purple Mountain sight spot Mei Linggong periphery includes many ages
Forest arbor more remote, such as Platanus acerifolia Platanus acerifolia, deodar Cedrus deodara, Quercus acutissima
The ancient and well-known trees such as Quercus acutissima are the component parts at former residence, and precious " historical relic living ".Thus building tool
It, need to be to folk culture, religious belief, humane feature, history relic, traditional tree during the forest landscape for having regional culture feature
The combined factors such as kind consider, analyze the humanistic environment of carrying.
According to characterization factor formula F 3, forest colorfulness is improved, can change the single green tint of large area to
Inflexible, unchanged, the unappealing visual experience of audience's bring, allows and visually brings impact, improves audience to forest
The degree of liking of landscape.Various plants flower, leaf all have color abundant in the forest community of the Purple Mountain, and ornamental value is high, such as can
See flower, littleleaf photinia root Photinia parvifolia for seeing leaf, match mountain plum Styrax confusus, rosaceae Rosaceae etc.
Most of color leafed plants；Folium Alniphylli fortumei Alniphyllum fortunei, Celtis julianae Celtis julianae, the coptis of considerable leaf
Wooden Pistacia chinensis, Chinese tallow tree Sapium sebiferum etc..But on-site inspection is, it was also found that most tradition coloured silk Ye Shu
Kind, such as Acer palmatum Acer palmatum, Chinese pistache Pistacia chinensis, Chinese tallow tree Sapium sebiferum, although
There is NATURAL DISTRIBUTION, but distribution is scattered about like the stars, and can not form strong aspect color landscape, therefore can suitably mend in some regions
It plants, to strengthen seasonal phenomena, constitutes strong visual Landscape.
According to characterization factor formula F 4, the public is more demanding to the property easy to identify of forest arbor, this is because in current city
City's plants landscape constantly it is newest ask different during, people are weaker for novel plant variety, adventitious plant cognitive power, vision
Preference is general, and the well known indigenous plant (features such as tree crown, trunk are easily identified) that the vision of people is easier quilt attracts.
Evergreen species tree Qinggang Cyclobalanopsis glauca, bitter sweet oak Castanopsis in the forest community of the Purple Mountain
Sclerophylla, Chinese photinia Photinia serrulata, Chinese ilex Ilex chinensis and deciduous species Quercus acutissima Quercus
Acutissima, cork oak Quercusvariabilis, white oak Quercus fabri, Chinese sweet gum Liquidambar
Formosana, yellow wingceltis Dalbergia hupeana, Chinese pistache Pistacia chinensis, Platycarya strobilacea Platycarya
Strobilacea, tilia miqueliana Tilia miqueliana are the known and favorite tree species of citizen.
According to characterization factor formula F 5, increases intervisibility degree in woods, can effectively improve the public for forest landscape landscape
Ornamental preference degree.The vision permeability between forest can be increased by artificial appropriate intermediate cutting, take different age cladding Mixed species
Forest landscape transformation of forest phase is carried out, implements " hayashishita is inserted green " engineering in hayashishita.Purple Mountain coniferous forest is in nineteen eighty-two insect infestation, pine
The plants such as tree are largely extremely busy, cause coniferous forest landscape large area ruined.Implement to remove the tree species that die of illness at that time, increase penetrating in woods
Property, and excellent cold-resistant evergreen broad-leaved landscape tree kind is selected to implement " hayashishita is inserted green " in hayashishita, these present plants have become hayashishita
Main associated species becomes the high forest landscape landscape of ornamental value.
It can be seen that the visual quality for influencing forest scenery resources is based primarily upon following 5 principal components: forest landscape is closely certainly
Right degree, forest landscape regional culture feature, forest colorfulness, arbor feature easy to identify and Lin Nei spatial impression.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, although referring to aforementioned reality
Applying example, invention is explained in detail, for those skilled in the art, still can be to aforementioned each implementation
Technical solution documented by example is modified or equivalent replacement of some of the technical features.It is all in essence of the invention
Within mind and principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of forest scenery resources visual quality evaluation method based on image and principal component, which is characterized in that including following
Step:
S1, forest scenery resources are collected in such a way that two kinds of equipment of unmanned plane and field camera record 4K image data
Image data；
S2, visual quality evaluation personnel carry out visual quality using image data of the Likert scale to forest scenery resources and comment
Valence；
S3, visual quality evaluation personnel visual quality impact factor of interest is obtained using questionnaire survey mode；
S4, using SPSS23.0 software, dimensionality reduction is carried out to impact factor with Principal Component Analysis, so that obtaining meeting the public needs
The visual signature factor for the forest scenery resources asked.
2. a kind of forest scenery resources visual quality evaluation method based on image and principal component according to claim 1,
It is characterized in that, the unmanned plane is 2 Pro unmanned plane of DJI-Mavic in the S1, the field camera is DJI-
The image for the forest scenery resources that Osmo pocket field camera, the unmanned plane and the field camera are recorded
Data are the experiment sample of 4K format；What the forest landscape of the forest scenery resources selected is area for 400 ㎡ (20m*20m)
Forest vegetation, the recording of image data is divided into two sections, respectively includes the simulated altitude angle viewing of unmanned plane recording
The scene for simulation people's angle viewing forest landscape that the scene and video camera of forest landscape are recorded.
3. a kind of forest scenery resources visual quality evaluation method based on image and principal component according to claim 1,
It is characterized in that, the visual quality evaluation personnel is no colour blindness, without the public of anomalous trichromatism, based on social in the S2
Sighting distance is unfolded the evaluation of forest scenery resources visual quality, the social sighting distance, that is, public outside woods 70-100m distance it is ornamental.
4. a kind of forest scenery resources visual quality evaluation method based on image and principal component according to claim 3,
It is characterized in that, the visual quality evaluation personnel only has the interior of audio-visual playback equipment, evaluation between individually being brought into one
Process includes following three step:
A, to visual quality evaluation personnel, this time the whole of evaluation content is introduced, and plays all image data, so as to
Visual quality evaluation personnel can quickly form the overall impression of forest landscape；
B, two sections of image datas for individually playing forest landscape, the visual quality watched for forest landscape make evaluation of estimate；Depending on
Feel that quality evaluation personnel use Likert scale when evaluating to forest landscape, " 5,3,1,0, -1, -3, -5 " is used to respectively correspond
" very high, high, higher, general, lower, low, very low " of visual quality.
C, evaluation is made for visual quality impact factor, opinion scale positions 1,2,3,4,5 Pyatyi score values.
5. a kind of forest scenery resources visual quality evaluation method based on image and principal component according to claim 4,
It is characterized in that, being handled for the evaluation that the visual quality is made using Spass23.0 software standardization, at the standardization
Manage formula are as follows:
Valueij=(Rij-Rj)/Sj；Valuei=∑ Zij/Nj；
Symbol meaning is as follows: ValueijIt is jth position visual quality evaluation personnel to the evaluation criterion value of forest landscape at i-th, Rij
It is jth position estimator to the evaluation of estimate of forest landscape at i-th；RjIt is jth position visual quality evaluation personnel to all forest landscapes
Evaluation average value, RjIt is poor for evaluation criterion of the jth position estimator to all forest landscapes；ValueiFor forest landscape at i-th
Evaluation criterion value, NjFor all estimator's numbers.
6. a kind of forest scenery resources visual quality evaluation method based on image and principal component according to claim 4,
It is characterized in that, the score value for the visual quality impact factor is influenced using the immediate score value of mean value as the visual quality
The scoring criteria of the factor, formula are as follows:
Fik=∑ Iijk/Njk
Symbol meaning is as follows: FikFor the average value of k visual quality impact factors of i-th piece of forest landscape, IijkFor jth position vision
Evaluation of estimate of the quality evaluation personnel to k visual quality impact factors of i-th piece of forest landscape, NjkFor all of k metrics evaluations
Estimator's number.
7. a kind of forest scenery resources visual quality evaluation method based on image and principal component according to claim 4,
It is characterized in that, the visual quality impact factor includes: ecological factor, aesthetic values, cultural words, Mental Origin, time
Five dimensions of history.
8. a kind of forest scenery resources visual quality evaluation method based on image and principal component according to claim 1,
It is characterized in that, in the S3, the questionnaire survey mode includes to age bracket, profession, educational background, local and every year going forest
Park number is the visual quality evaluation personnel of background.
9. a kind of forest scenery resources visual quality evaluation method based on image and principal component according to claim 1,
It is characterized in that, whether analyzing visual quality evaluation personnel sample size using the normalized form that Scheaffer is proposed in the S3
With enough representativenesses:
Wherein n is visual quality evaluation personnel quantity, and N is the questionnaire survey size of population that statistics represents, and δ is sample error, works as δ
When taking 0.05, confidence coefficient >=95%, the Population that the statistics of n >=400 represents at this time；
Utilize the confidence level of the side reaction coefficient verifying questionnaire proposed by Cronbach, calculation formula are as follows:
K indicates the forest landscape number of pictures of questionnaire, SiIndicate the variance of respondent, Sx2What is indicated is all investigation pair
The variance of elephant.
10. a kind of forest scenery resources visual quality evaluation method based on image and principal component according to claim 1,
It is characterized in that, in the S4, the visual signature factor of the forest scenery resources includes forest landscape Nature closeness feature, gloomy
Woods landscape regional culture feature, forest colorfulness feature, arbor feature easy to identify and Lin Nei spatial impression feature；The forest
The factor of landscape Nature closeness feature includes: composition TCS, biotype composition LC, the vertical hierachy number SS of group and level pair
Than degree SC；The factor of the forest landscape regional culture feature includes: indigenous plant amount NPP, folk custom culture connotation degree CFC, drink
Diet culture intension degree FCC, religious belief cultural connotation degree CRB, Farming Culture intension degree CFC2 and time history TH；It is described gloomy
The factor of woods colorfulness feature includes aspect tree species ratio PST, pattern richness VDC, leaf color richness CR, color contrast
Spend CC and plant color color CI；The factor of the arbor feature easy to identify include trunk display degree VS, crown identification degree ITC and
Group kink characteristics CS；The factor of spatial impression feature includes canopy density CD and intervisibility degree DV in the woods.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910726844.7A CN110458088A (en) | 2019-08-07 | 2019-08-07 | A kind of forest scenery resources visual quality evaluation method based on image and principal component |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910726844.7A CN110458088A (en) | 2019-08-07 | 2019-08-07 | A kind of forest scenery resources visual quality evaluation method based on image and principal component |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110458088A true CN110458088A (en) | 2019-11-15 |
Family
ID=68485416
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910726844.7A Pending CN110458088A (en) | 2019-08-07 | 2019-08-07 | A kind of forest scenery resources visual quality evaluation method based on image and principal component |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110458088A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111444774A (en) * | 2020-02-26 | 2020-07-24 | 山西林业职业技术学院 | Forest resource monitoring method based on unmanned aerial vehicle aerial survey technology |
| CN113159377A (en) * | 2021-03-12 | 2021-07-23 | 江苏唱游数据技术有限公司 | Scenic spot smoothness prediction model method based on multi-factor aggregation model |
| CN113742918A (en) * | 2021-09-06 | 2021-12-03 | 中国科学院地理科学与资源研究所 | Method for evaluating recreation bearing capacity of national park based on system dynamics model |
| CN114298415A (en) * | 2021-12-29 | 2022-04-08 | 中国科学院地理科学与资源研究所 | Method and device for predicting pleasure degree of geographic environment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105069268A (en) * | 2015-09-11 | 2015-11-18 | 长江大学 | Landscape visual impact comprehensive evaluation method based on GIS |
| CN108846870A (en) * | 2018-06-12 | 2018-11-20 | 南京林业大学 | Scenic forest landscape aspect color aesthetics method |
| CN109146245A (en) * | 2018-07-17 | 2019-01-04 | 南京林业大学 | A kind of factor-analysis approach that the Garden Plant Landscape based on the elderly visual angle is built |
| CN109345314A (en) * | 2018-10-16 | 2019-02-15 | 湖南城市学院 | A kind of road landscape visual evaluation analysis method |
-
2019
- 2019-08-07 CN CN201910726844.7A patent/CN110458088A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105069268A (en) * | 2015-09-11 | 2015-11-18 | 长江大学 | Landscape visual impact comprehensive evaluation method based on GIS |
| CN108846870A (en) * | 2018-06-12 | 2018-11-20 | 南京林业大学 | Scenic forest landscape aspect color aesthetics method |
| CN109146245A (en) * | 2018-07-17 | 2019-01-04 | 南京林业大学 | A kind of factor-analysis approach that the Garden Plant Landscape based on the elderly visual angle is built |
| CN109345314A (en) * | 2018-10-16 | 2019-02-15 | 湖南城市学院 | A kind of road landscape visual evaluation analysis method |
Non-Patent Citations (2)
| Title |
|---|
| 孙姝亭: "《北京市居民小区景观林景观质量评价研究》", 《中国优秀硕士学位论文全文数据库 农业科技辑》 * |
| 杨阳 等: "《基于公众感知的牛首山风景区人工植物景观视觉评价》", 《山东农业大学学报（自然科学版）》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111444774A (en) * | 2020-02-26 | 2020-07-24 | 山西林业职业技术学院 | Forest resource monitoring method based on unmanned aerial vehicle aerial survey technology |
| CN113159377A (en) * | 2021-03-12 | 2021-07-23 | 江苏唱游数据技术有限公司 | Scenic spot smoothness prediction model method based on multi-factor aggregation model |
| CN113159377B (en) * | 2021-03-12 | 2024-03-12 | 江苏唱游数据技术有限公司 | Scenic spot smooth flow prediction model method based on multi-factor aggregation model |
| CN113742918A (en) * | 2021-09-06 | 2021-12-03 | 中国科学院地理科学与资源研究所 | Method for evaluating recreation bearing capacity of national park based on system dynamics model |
| CN113742918B (en) * | 2021-09-06 | 2023-10-20 | 中国科学院地理科学与资源研究所 | Method for evaluating national park recreation bearing capacity based on system dynamics model |
| CN114298415A (en) * | 2021-12-29 | 2022-04-08 | 中国科学院地理科学与资源研究所 | Method and device for predicting pleasure degree of geographic environment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110458088A (en) | A kind of forest scenery resources visual quality evaluation method based on image and principal component | |
| Horn | The adaptive geometry of trees | |
| Rudolph | Gnosis: The nature and history of Gnosticism | |
| Li et al. | Humanization of nature: Testing the influences of urban park characteristics and psychological factors on collegers’ perceived restoration | |
| Mahidin et al. | Understanding children preferences of natural environment as a start for environmental sustainability | |
| Huang et al. | An annales school-based serious game creation framework for taiwanese indigenous cultural heritage | |
| Nelson et al. | Perception of tree canopy | |
| CN108846870A (en) | Scenic forest landscape aspect color aesthetics method | |
| Deng et al. | Dynamic response of the scenic beauty value of different forests to various thinning intensities in central eastern China | |
| Tarakçı Eren et al. | Determination of the visual preference levels and perceptual differences in the appearance of certain taxa in different seasons | |
| Slater | Contested terrain: Differing interpretations of Co. Wicklow's landscape | |
| De Haldat Du Lys et al. | If self‐shading is so bad, why is there so much? Short shoots reconcile costs and benefits | |
| Kuper | Preference, complexity, and color information entropy values for visual depictions of plant and vegetative growth | |
| Serra et al. | Improvement of the integration of visually impacting architectures in historical urban scene, an application of semantic differencial method | |
| Eroglu et al. | Ecological and Visual Planting Design Analysis. A Case Study of two Parks in Amsterdam | |
| Karmanov et al. | Evaluations of design gardens by students of landscape architecture and non-design students: A comparative study | |
| Yang et al. | The aesthetic evaluation of forest landscape in four urban forest parks: A case study of Nanjing hilly areas, China | |
| Alon-Mozes | Gardens and the emergence of culture: The Meir Garden: The first Hebrew garden for the first Hebrew city—Tel Aviv | |
| Galev | Positive and unfavourable visual effects of the forest landscapes | |
| ÖZHANCI | Comparative perception analysis in special-qualified heritage landscapes | |
| Kigozi et al. | Exposing Uganda’s Cultural Mosaic: An examination of the Endingidi’s origins, educational and aesthetic dimensions | |
| Wheeler | Chercheur de Lumière: The Art of Jean-Louis Rouméguère (1863-1925) | |
| Meng et al. | Study on Five-sense Landscape of Chu Ci Plants Based on NVivo | |
| Erhabor et al. | Ecological Study of Indigenous Tree Population and Potentials in Urban Green Space: A Case Study of Benin Teaching Hospitals, Benin-City, Edo State, Nigeria | |
| Akhir et al. | Planting Design Expression on Students Perceiving Campus Landscape Quality |
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: 20191115 |