CN109087321A - A kind of profile testing method based on visual pathway multi-layer inhibition zone synergistic effect - Google Patents
A kind of profile testing method based on visual pathway multi-layer inhibition zone synergistic effect Download PDFInfo
- Publication number
- CN109087321A CN109087321A CN201810575947.3A CN201810575947A CN109087321A CN 109087321 A CN109087321 A CN 109087321A CN 201810575947 A CN201810575947 A CN 201810575947A CN 109087321 A CN109087321 A CN 109087321A
- Authority
- CN
- China
- Prior art keywords
- neuron
- response
- inhibition zone
- visual cortex
- primary visual
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
- G06N3/061—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using biological neurons, e.g. biological neurons connected to an integrated circuit
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- General Physics & Mathematics (AREA)
- Neurology (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Microelectronics & Electronic Packaging (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
- Eye Examination Apparatus (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to a kind of profile testing methods based on visual pathway multi-layer inhibition zone synergistic effect.The present invention constructs the foreign journals neuron arrays of orientation-sensitive first, and the contour direction of record foreign journals selection calculates the response of foreign journals neuron；Vision input with distance for the classical receptive field of the external knee somatic nerves member of factor and inhibition zone carries out quantification treatment, using the two difference as inhibition zone significant response, normalizes through active rectification and power exponent, obtains synergic parameters；The vision input of primary visual cortex is obtained by the response of local window Weighted Fusion foreign journals neuron, detects accurate profile response；The response of primary visual cortex neuron inhibition zone is calculated, final profile response is obtained.The present invention considers the fining testing mechanism step by step when different levels neuron perceived direction in visual pathway, while simulating the synergistic effect of inhibition zone between different levels, can effectively improve the contour detecting performance of natural image.
Description
Technical field
The invention belongs to field of machine vision, relate generally to a kind of based on visual pathway multi-layer inhibition zone synergistic effect
Profile testing method.
Background technique
Contour detecting is one of image understanding or the important link of visual analysis mid-early stage, and the contour feature of acquisition will have
Vision Key detail after effect expression de-redundancy.The difficult point of contour detecting task was mainly manifested in detection and owed detection two
Aspect, the former is due to the interference by false contourings such as texture noises, and the latter is then since image is in the distribution of silhouette contrast degree
Otherness.Traditional profile testing method is based primarily upon the space transition of image information, therefore takes with difference or morphology
Mathematical measure based on equal operations, usually can obtain good performance in the preferable situation of picture quality.But tradition side
Method ignores the biological neural mechanism in visual perception completely, therefore when in face of the complex task such as multistage contour detecting,
It is difficult to meet the requirement of detection performance.Although currently based on the profile testing method mould to a certain extent of biological vision mechanism
Intend the ability that biological vision extracts image outline feature, but they are when simulating visual pathway visual information stream process process,
The classical receptive field of some level neuron or inhibition zone itself in visual pathway more are focused on, and is had ignored in visual pathway not
It acts synergistically with inhibition zone possessed between level, this act synergistically changes the orphan of each level of human perceptual model
Vertical property, can go to give full play to the key player that each level is played the part of in perception with the characteristics of overall coordination.
Summary of the invention
In view of the deficiencies of the prior art, the present invention proposes a kind of based on visual pathway multi-layer inhibition zone synergistic effect
Profile testing method.
This patent is considered: (1) primary visual cortex neuron has direction selection characteristic, traditional detection to visual stimuli
Method is normally set up the angle of several discretizations, and the angle for response extreme value occur then is determined as contour direction, is examined at this time
The performance of survey having significant consumption detection system by discretization angle closely related with the degree of angular discretization, intensive
Resource is limited, the discretization angle of dispersion will seriously affect the accuracy of detection again.(2) conventional method is in the primary view skin of simulation
When the direction selection characteristic of layer neuron, comprehensively considering for angle and scale is generally only carried out in this level, it is not intended that preceding
Grade link is to the synergistic effect of primary visual cortex neuron, and the relevance between this level will be helpful to from local detail
It is transitioned into integrally-built description.(3) conventional method portrays the single level of primary visual cortex usually using difference of Gaussian model
Interior inhibition zone properties, and have ignored the optic nerve mechanism that multi-layer inhibition zone acts synergistically in visual pathway.This collaboration
Effect will make the transmitting of visual information stream in visual pathway and processing constitute one it is effectively whole, be beneficial to contour detecting from
Part and global visual angle expansion fusion.
Therefore, step by step fining detection machine of the present invention from visual pathway when different levels neuron perceived direction is produced
Hair, while the synergistic effect of inhibition zone between different levels is simulated, it proposes a kind of based on visual pathway multi-layer inhibition zone synergistic effect
Profile testing method, give full play to above-mentioned visual characteristic in the important function of contour detecting.
The invention mainly comprises the following steps:
Step (1) constructs and the one-to-one foreign journals neuron arrays of pixel first；Then simulation foreign journals nerve
The direction selection mechanism of member, sets k discrete direction to be detected, for foreign journals corresponding to each pixel I (x, y)
Neuron calculates the accounting of each directional response intensity and direction response intensity and all directional response intensity summations.
If the maximum value of accounting is more than threshold value, direction corresponding to this maximum value is recorded, optimal response direction ID is considered as1(x,y)；
If all accountings are not up to threshold value, direction corresponding to preceding two big numerical value of accounting, i.e. optimal response direction are recorded simultaneously
Direction is responded with suboptimum, is denoted as ID respectively1(x, y) and ID2(x,y).Finally using the maximum of directional response as the foreign journals
The response of neuron.
Step (2) calculates foreign journals-primary visual cortex inhibition zone synergic parameters wLGN(x,y)。
The vision input that foreign journals neuron classics receptive field is obtained first with Gaussian function, utilizes double gauss difference letter
Number obtains the vision input of foreign journals neuron inhibition zone, and is based respectively on distance factor and quantifies to two vision inputs
Processing；Then using the difference of quantified treated two visions input as the significant response of foreign journals neuron inhibition zone,
And dynamic halfwave rectifier is carried out to significant response.Power exponent normalization finally is carried out to the significant response after dynamic halfwave rectifier；
Power exponent normalization result is set to adjust the synergic parameters w of primary visual cortex neuron inhibition zone responseLGN(x,
y)。
Step (3) building, with the primary visual cortex neuron arrays of size, determines primary view with foreign journals neuron arrays
The vision of each neuron of cortex inputs U (x, y).
Receptive field of the local window as some neuron of primary visual cortex is constructed first, then calculates receptive field
The Synaptic junction weight of interior all foreign journals neurons and the primary visual cortex neuron finally merges all outer knees in window
The response of somatic nerves member obtains vision input U (x, y) of the primary visual cortex neuron.
Step (4) realizes that the direction of primary visual cortex neuron refines testing mechanism.
If foreign journals only have recorded ID in step (1)1(x, y) is then spaced in optimal response based on fixed discrete angular
Direction bilateral symmetry respectively sets q discrete direction；If foreign journals have recorded ID simultaneously1(x, y) and ID2(x, y), then optimal and
Suboptimum response sets p discrete direction between direction altogether.For above-mentioned 2q or p discrete direction, detection is primary respectively regards skin
Directional response of the layer neuron when vision inputs U (x, y).The essence that directional response maximum value is detected as primary visual cortex
Quasi- profile responds EV1(x,y)。
Step (5) calculates primary visual cortex inhibition zone response Inh (x, y) of fusion space scale factor first.Then sharp
The synergic parameters w obtained with step (2)LGNInh (x, y) is adjusted in (x, y), finally is used to inhibit by adjusted result
Accurate profile responds EV1(x, y), and the response after inhibition is responded into E (x, y) as final profile.
The device have the advantages that are as follows:
1. simulating the fining testing mechanism step by step when multi-layer neuron detection contour direction, detection profile is reduced
The coupling in direction and angular discretization degree, while the continuity in direction is met to a certain extent, make full use of detection
The limited resource of system, detects contour direction faster more quasi-ly.
2. simulating the synergistic effect mechanism of multi-layer inhibition zone, building foreign journals-primary visual cortex inhibition zone collaboration is made
With model, synergic parameters are extracted in the significant response based on foreign journals inhibition zone, to the inhibition zone intensity of primary visual cortex into
Row is adjusted, weaken between profile from inhibiting, going unless profile texture and pseudo-edge protect actual profile simultaneously, more
Meet the information processing manner that inhibition zone acts synergistically in visual pathway.
Detailed description of the invention
Fig. 1 is flow chart of the invention.
Specific embodiment
In conjunction with attached drawing 1, the specific implementation steps of the present invention are as follows:
Step (1) constructs and the one-to-one foreign journals neuron arrays of pixel first.Then simulation foreign journals nerve
The direction selection mechanism of member sets a discrete direction of k (being defaulted as 6) to be detected as shown in formula (1).Then it is directed to each picture
Foreign journals neuron corresponding to vegetarian refreshments I (x, y) obtains specific direction θ as shown in formula (2)iResponse intensity eLGN(x,y；θi,
σl).Calculate simultaneously the response intensity of the direction with directive response intensity summation accounting d (x, y；θi), such as formula (3) institute
Show.
Wherein * is convolution symbol, and I (x, y) is input picture,
σlIt is the size of foreign journals neuron classics receptive field, default is set as 2.If the maximum value d of accountingmax(x,y；θi) it is more than threshold value
(being defaulted as 0.5) then records direction corresponding to this maximum value, is considered as optimal response direction ID1(x,y)；If all accountings are equal
Not up to threshold value then records direction corresponding to preceding two big numerical value of accounting, i.e. optimal response direction and suboptimum responder simultaneously
To being denoted as ID respectively1(x, y) and ID2(x,y).Finally using the maximum of directional response as corresponding to pixel I (x, y)
Foreign journals neuron responds ELGN(x, y), as shown in formula (4).
ELGN(x, y)=max { eLGN(x,y；θi,σl) | i=0,1 ... k-1 } (4)
Step (2) calculates foreign journals-primary visual cortex inhibition zone synergic parameters wLGN(x,y).First with height
This function obtains the vision input of foreign journals neuron classics receptive field, obtains foreign journals neuron using double gauss difference function
The vision of inhibition zone inputs, and is based respectively on distance factor and carries out quantification treatment to two vision inputs.WithIt is the vision input after classical receptive field and inhibition zone quantification treatment respectively, defines as shown in formula (5)~(8).
DoG+(m,n；σl,ρl)=max { gaus (m, n, ρlσl)-gaus(m,n,σl),0} (8)
Wherein ρlFor the ratio of foreign journals neuron inhibition zone and classical receptive field size, default, which is set as 4, m, n, indicates warp
Offset in allusion quotation receptive field between vision input I (x+m, y+n) and I (x, y).Then willDifferenceIt is moved as the significant response of foreign journals neuron inhibition zone, and to Δ φ (x, y)
State halfwave rectifier will be less than threshold value that is, using the average value mean of Δ φ (x, y) as threshold value0 is set, threshold value is not less than
Remain unchanged, as shown in formula (9).
Power exponent normalization finally is carried out to the Δ φ (x, y) after dynamic halfwave rectifier, and primary visual cortex mind will be adjusted
The synergic parameters w responded through first inhibition zoneLGN(x, y) setting normalizes as a result, as shown in formula (10) thus.
Wherein, δ is inhibition zone significant response adjustment parameter, default value 2.
Step (3) building, with the primary visual cortex neuron arrays of size, determines primary view with foreign journals neuron arrays
The vision of each neuron of cortex inputs U (x, y).A local window W is constructed firstxyAs some nerve of primary visual cortex
Then the receptive field of member calculates the Synaptic junction of all foreign journals neurons and the primary visual cortex neuron in receptive field and weighs
Value w (xt,yt), as shown in formula (11):
Wherein r is local window radius (being defaulted as 3), (xt,yt) be expressed as t-th in window centered on (x, y)
Foreign journals neuron finally merges the response of all foreign journals neurons in window, obtains the view of the primary visual cortex neuron
Input U (x, y) is felt, as shown in formula (12).
Wherein μ is that all foreign journals neurons respond summation in window.
Step (4) realizes that the direction of primary visual cortex neuron refines testing mechanism.If foreign journals are only in step (1)
Have recorded ID1(x, y), then it is discrete in each setting of optimal response direction bilateral symmetry q based on fixed discrete angular interval s
Direction, as shown in formula (13)；If foreign journals have recorded ID simultaneously1(x, y) and ID2(x, y) then responds direction in optimal and suboptimum
Between equidistantly set total p discrete direction, as shown in formula (14).Default sets s=5 °, p=3, q=6.
θh=ID1(x, y)+(h-q) × s, h=0,1 ... 2q-1 (13)
For above-mentioned 2q or p discrete direction, primary visual cortex neuron is detected respectively in vision and inputs U (x, y),
Direction is θhWhen directional response eV1(x,y；θh,σv), as shown in formula (15).
WhereinσvIt is that primary visual cortex neuron is classical
The size of receptive field, default value are set as 4.The accurate profile response that directional response maximum value is detected as primary visual cortex
EV1(x,y).As shown in formula (16).
EV1(x, y)=max { eV1(x,y；θh,σv) | h=0,1 ... 2q-1OR h=0,1 ... p-1 } (16)
Step (5) calculates primary visual cortex neuron inhibition zone response Inh (x, y) of fusion space scale factor first,
As shown in formula (17), (18).
Wherein, | DoG+(x,y；σv,ρv) | it is DoG+(x,y；σv,ρv) L1 norm, ρvIt is the suppression of primary visual cortex neuron
The ratio in area processed and classical receptive field, default value 4, the synergic parameters w then obtained using step (2)LGN(x, y) is right
Inh (x, y) is adjusted.Finally adjusted result is used to accurate profile be inhibited to respond EV1(x, y), and by the response after inhibition
E (x, y) is responded as final profile.As shown in formula (19).
E (x, y)=EV1(x,y)-α·wLGN(x,y)·Inh(x,y) (19)
Wherein, α is the parameter for adjusting inhibition zone entirety inhibition strength, is defaulted as 1.0.
Claims (1)
1. a kind of profile testing method based on visual pathway multi-layer inhibition zone synergistic effect, which is characterized in that this method is specific
The following steps are included:
Step (1) constructs and the one-to-one foreign journals neuron arrays of pixel first；Then foreign journals neuron is simulated
Direction selection mechanism sets k discrete direction to be detected, for the nerve of foreign journals corresponding to each pixel I (x, y)
Member calculates the accounting of each directional response intensity and direction response intensity and all directional response intensity summations；If accounting
Maximum value be more than threshold value, then record direction corresponding to this maximum value, be considered as optimal response direction ID1(x,y)；If all account for
Than being not up to threshold value, then direction corresponding to preceding two big numerical value of accounting is recorded simultaneously, i.e. optimal response direction and suboptimum is rung
Direction is answered, is denoted as ID respectively1(x, y) and ID2(x,y)；Finally using the maximum of directional response as the sound of the foreign journals neuron
It answers；
Step (2) calculates foreign journals-primary visual cortex inhibition zone synergic parameters wLGN(x,y)；
The vision input that foreign journals neuron classics receptive field is obtained first with Gaussian function, is obtained using double gauss difference function
The vision of foreign journals neuron inhibition zone inputs, and is based respectively on distance factor and carries out quantification treatment to two vision inputs；So
Afterwards using the difference of quantified treated two visions input as the significant response of foreign journals neuron inhibition zone, and to effective
Response carries out dynamic halfwave rectifier；Power exponent normalization finally is carried out to the significant response after dynamic halfwave rectifier；By power exponent
Normalization result is set as adjusting the synergic parameters w of primary visual cortex neuron inhibition zone responseLGN(x,y)；
Step (3) building, with the primary visual cortex neuron arrays of size, determines primary visual cortex with foreign journals neuron arrays
The vision of each neuron inputs U (x, y)；
Receptive field of the local window as some neuron of primary visual cortex is constructed first, then calculates in receptive field and owns
The Synaptic junction weight of foreign journals neuron and the primary visual cortex neuron finally merges all foreign journals neurons in window
Response, obtain vision input U (x, y) of the primary visual cortex neuron；
Step (4) realizes that the direction of primary visual cortex neuron refines testing mechanism；
If foreign journals only have recorded ID in step (1)1(x, y) is then spaced in an optimal response direction left side based on fixed discrete angular
Right symmetrical respectively q discrete direction of setting；If foreign journals have recorded ID simultaneously1(x, y) and ID2(x, y) is then rung in optimal and suboptimum
It answers and sets p discrete direction between direction altogether；For above-mentioned 2q or p discrete direction, primary visual cortex nerve is detected respectively
Directional response of the member when vision inputs U (x, y)；The accurate profile that directional response maximum value is detected as primary visual cortex is rung
Answer EV1(x,y)；
Step (5) calculates primary visual cortex inhibition zone response Inh (x, y) of fusion space scale factor first；Then step is utilized
Suddenly the synergic parameters w that (2) obtainLGNInh (x, y) is adjusted in (x, y), is finally used to adjusted result inhibit accurate
Profile responds EV1(x, y), and the response after inhibition is responded into E (x, y) as final profile.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810575947.3A CN109087321B (en) | 2018-06-06 | 2018-06-06 | Profile detection method based on visual path multi-level inhibition area synergistic effect |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810575947.3A CN109087321B (en) | 2018-06-06 | 2018-06-06 | Profile detection method based on visual path multi-level inhibition area synergistic effect |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109087321A true CN109087321A (en) | 2018-12-25 |
| CN109087321B CN109087321B (en) | 2021-08-10 |
Family
ID=64839425
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810575947.3A Active CN109087321B (en) | 2018-06-06 | 2018-06-06 | Profile detection method based on visual path multi-level inhibition area synergistic effect |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109087321B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109978898A (en) * | 2019-02-01 | 2019-07-05 | 广西科技大学 | Profile testing method based on vector field energy balane |
| CN111402285A (en) * | 2020-01-16 | 2020-07-10 | 杭州电子科技大学 | Contour detection method based on visual mechanism dark edge enhancement |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102201120A (en) * | 2011-06-17 | 2011-09-28 | 电子科技大学 | Multifeature-based target object contour detection method |
| CN103473759A (en) * | 2013-06-24 | 2013-12-25 | 南京理工大学 | Low-light-level image significant contour extraction method of WKPCA homogeneity degree correction nCRF inhibition |
| CN104778696A (en) * | 2015-04-13 | 2015-07-15 | 杭州电子科技大学 | Image edge grading-detection method based on visual pathway orientation sensitivity |
| CN106127740A (en) * | 2016-06-16 | 2016-11-16 | 杭州电子科技大学 | A kind of profile testing method based on the association of visual pathway many orientation of sensory field |
-
2018
- 2018-06-06 CN CN201810575947.3A patent/CN109087321B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102201120A (en) * | 2011-06-17 | 2011-09-28 | 电子科技大学 | Multifeature-based target object contour detection method |
| CN103473759A (en) * | 2013-06-24 | 2013-12-25 | 南京理工大学 | Low-light-level image significant contour extraction method of WKPCA homogeneity degree correction nCRF inhibition |
| CN104778696A (en) * | 2015-04-13 | 2015-07-15 | 杭州电子科技大学 | Image edge grading-detection method based on visual pathway orientation sensitivity |
| CN106127740A (en) * | 2016-06-16 | 2016-11-16 | 杭州电子科技大学 | A kind of profile testing method based on the association of visual pathway many orientation of sensory field |
Non-Patent Citations (3)
| Title |
|---|
| KAI-FU YANG等: "Multifeature-Based Surround Inhibition Improves Contour Detection in Natural Images", 《IEEE TRANSACTIONS ON IMAGE PROCESSING》 * |
| 杨开富: "基于多视觉特征的非经典感受野模型及应用研究", 《中国优秀硕士学位论文全文数据库》 * |
| 王典等: "基于突触连接视通路方位敏感的图像分级边缘检测方法", 《中国生物医学工程学报》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109978898A (en) * | 2019-02-01 | 2019-07-05 | 广西科技大学 | Profile testing method based on vector field energy balane |
| CN109978898B (en) * | 2019-02-01 | 2023-07-18 | 广西科技大学 | Contour detection method based on vector field energy calculation |
| CN111402285A (en) * | 2020-01-16 | 2020-07-10 | 杭州电子科技大学 | Contour detection method based on visual mechanism dark edge enhancement |
| CN111402285B (en) * | 2020-01-16 | 2023-08-08 | 杭州电子科技大学 | Contour detection method based on visual mechanism dark edge enhancement |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109087321B (en) | 2021-08-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106228547B (en) | A kind of profile and border detection algorithm of view-based access control model color theory and homogeneity inhibition | |
| Komori et al. | Effect of averageness and sexual dimorphism on the judgment of facial attractiveness | |
| CN109489576B (en) | Contour detection method based on primary visual path calculation model | |
| CN110706826B (en) | Non-contact real-time multi-person heart rate and blood pressure measuring method based on video image | |
| CN111402285B (en) | Contour detection method based on visual mechanism dark edge enhancement | |
| CN112288010B (en) | Finger vein image quality evaluation method based on network learning | |
| CN101546430A (en) | Edge extracting method based on simplified pulse coupled neural network | |
| CN110046675A (en) | A kind of the exercise ability of lower limbs appraisal procedure based on improved convolutional neural networks | |
| CN109544523B (en) | Method and device for evaluating quality of face image based on multi-attribute face comparison | |
| CN106127740B (en) | One kind being based on the associated profile testing method of the more orientation of sensory field of visual pathway | |
| CN109087321A (en) | A kind of profile testing method based on visual pathway multi-layer inhibition zone synergistic effect | |
| CN104778696B (en) | A kind of image border hierarchical detection method based on visual pathway azimuth sensitivity | |
| CN112472048A (en) | Neural network structure for pulse condition recognition of cardiovascular disease patient | |
| CN113850013B (en) | Ship radiation noise classification method | |
| CN104881851A (en) | Adaptive image target enhancement method based on difference of Gaussian model | |
| Lei et al. | Finger-vein image enhancement based on pulse coupled neural network | |
| CN111179294B (en) | Bionic type contour detection method based on X, Y parallel visual channel response | |
| CN102222231B (en) | Visual attention information computing device based on guidance of dorsal pathway and processing method thereof | |
| CN108805879A (en) | A kind of image partition method based on Spiking neural networks | |
| CN109558797A (en) | The method that center of gravity Area Model under view-based access control model stimulation distinguishes total balance of the body obstacle | |
| CN106709480B (en) | Intersected human face recognition methods based on weighed intensities PCNN models | |
| CN107392912A (en) | Image partition method based on pseudo-color coding and DISCOV codings | |
| CN108492275A (en) | Based on deep neural network without with reference to stereo image quality evaluation method | |
| CN103955695B (en) | The method that computer is based on human eye state in gray level co-occurrence matrixes energy variation Intelligent Recognition video | |
| Anguraj et al. | Evaluation and severity classification of facial paralysis using salient point selection algorithm |
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 |