CN110062502A - A kind of online predicting residual useful life of LED illumination lamp based on machine vision and reliability estimation method - Google Patents
A kind of online predicting residual useful life of LED illumination lamp based on machine vision and reliability estimation method Download PDFInfo
- Publication number
- CN110062502A CN110062502A CN201910322099.XA CN201910322099A CN110062502A CN 110062502 A CN110062502 A CN 110062502A CN 201910322099 A CN201910322099 A CN 201910322099A CN 110062502 A CN110062502 A CN 110062502A
- Authority
- CN
- China
- Prior art keywords
- lamps
- lanterns
- image
- reliability
- brightness
- 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
-
- H—ELECTRICITY
- H05—ELECTRIC TECHNIQUES NOT OTHERWISE PROVIDED FOR
- H05B—ELECTRIC HEATING; ELECTRIC LIGHT SOURCES NOT OTHERWISE PROVIDED FOR; CIRCUIT ARRANGEMENTS FOR ELECTRIC LIGHT SOURCES, IN GENERAL
- H05B45/00—Circuit arrangements for operating light-emitting diodes [LED]
- H05B45/50—Circuit arrangements for operating light-emitting diodes [LED] responsive to malfunctions or undesirable behaviour of LEDs; responsive to LED life; Protective circuits
- H05B45/58—Circuit arrangements for operating light-emitting diodes [LED] responsive to malfunctions or undesirable behaviour of LEDs; responsive to LED life; Protective circuits involving end of life detection of LEDs
Landscapes
- Spectrometry And Color Measurement (AREA)
- Circuit Arrangement For Electric Light Sources In General (AREA)
Abstract
The invention discloses a kind of LED illumination lamp predicting residual useful life and reliability estimation method based on machine vision supports the online life prediction of LED illumination lamp and reliability assessment, belongs to Reliability Engineering field.Steps are as follows: the arrangement of 1. image information collecting equipment, according to lamp applications scene, according to standard arrangement image capture device as defined in this method.2. image information pre-processes, the image collected is performed corresponding processing, obtains qualified image information.3. brightness/chroma information identifies, identifies lamp brightness/coloration information in image information, determine region-of-interest and calculate its norm.4. establishing and lamps and lanterns luminous flux degeneration/color coordinate drift model being trained to establish luminous flux/chromaticity coordinates degradation model and reliability model respectively according to the brightness/chroma norm of identification arrived.5. carrying out tested lamps and lanterns predicting residual useful life and reliability assessment, lamps and lanterns predicting residual useful life and reliability assessment are carried out using the trained model arrived.
Description
Technical field
The present invention provides a kind of LED illumination lamp predicting residual useful life and reliability estimation method based on machine vision,
It supports the online life prediction of LED illumination lamp and reliability assessment, belongs to Reliability Engineering field.
Background technique
White light large-power light-emitting diodes (light-emitting diodes, LED) gradually substitution conventional light source at
For the lighting source of mainstream.It is potassium nitride (GaN) blue-light LED chip in the white light LEDs principle of luminosity that engineering field largely uses
Excitated fluorescent powder (yellow, orange etc.), to emit white light, main failure mode includes that luminous flux degeneration is inclined with chromaticity coordinates
It moves.Since LED product generally has the characteristics that long service life, high reliablity.To guarantee research and development of products and industrial production efficiency,
LED manufacturer generallys use use of the degraded data obtained in reliability and life test (or accelerated test) to product batch
Service life is assessed with reliability, and is made nominal information and issued together with product.But due to the hair of LED illumination lamp
Light principle, technology, use environment are complex, cause the dispersibility of lamp life larger, the nominal life that producer announces
The real surplus life-span and reliability of the LED lamp in use are not can accurately reflect generally with reliability information.And for shining
The brightness in Mingguang City source and color have the occasion such as exhibition center, medical space, job that requires special skills place etc. being strict with, accurate evaluation
The remaining life of LED illumination lamp in use can formulate reasonable maintenance support strategy with reliability information for user and provide
Effectively support.For the predicting residual useful life and reliability of the LED illumination lamp in assessment use, need to obtain real-time light source letter
Such as luminous flux, chromaticity coordinates are ceased, but information above is difficult to measure by lamps and lanterns own system.How to exist in LED illumination light source
Its light source information is measured under line working condition, and accurate predicting residual useful life and reliability assessment are carried out to it, is headlamp
Has the challenge in Reliability Engineering field.
Machine vision technique includes three image procossing, image recognition and image understanding links.Have benefited from advanced identification
The development of algorithm such as artificial neural network, support vector machines, Hidden Markov Model etc., machine vision technique are swift and violent in recent years
Development, image recognition accuracy, image analysis capabilities, image understanding efficiency are improved.So that machine vision technique is all
It is multi-field to have obtained sufficient application.
Summary of the invention
It is an object of the present invention to provide a kind of, and the LED illumination lamp predicting residual useful life based on machine vision is commented with reliability
Estimate on-line testing method.This method can detect the light source letter of illuminator by machine vision technique under noncontact condition
Breath, so that the real-time remaining life and reliability for accurate evaluation LED illumination lamp provide support, mainly comprises the steps of.
Step 1: the arrangement of image information collecting equipment and optical screen
Consider the different mounting means and usage scenario of illuminator, according to the focal length of image capture device and its with illumination
The relative position of lamps and lanterns, is arranged image capture device, guarantees that it can effectively collect lamps and lanterns and be incident upon on optical screen
Image.RGB triple channel still image of the resolution ratio greater than 72ppi that image capture device need to acquire Fixed Time Interval is believed
Breath.Optical screen is the flat surface that can be connected to the more complete picture of lamps and lanterns that distance is tested the nearest position of lamps and lanterns and angle is fixed.
According to the actual situation, optical screen can be indoor decoration plane, fixed article plane etc., can also install fixed reflective surface additional on lamps and lanterns
As optical screen.
Step 2: image information pretreatment
Based on machine vision technique, still image size, the contrast, luminance information of the lamps and lanterns image acquired are adjusted,
Sharpening appropriate, filtering processing are carried out, guarantees that lamps and lanterns image contour is clear, comparison of light and shade is distinct.To what is acquired in different moments
The image of same tested lamps and lanterns carries out completely the same adjustment and processing.Base is established in the position of the lamps and lanterns image obtained according to identification
On schedule, the image of the same tested lamps and lanterns of alignment different moments acquisition.After the completion of above step, treated single-pass is generated
The object that road (gray scale) picture is identified as luminous flux information retains the object that triple channel picture is identified as color coordinate information.
Step 3: brightness/chroma information identification
The highest region of lamps and lanterns image brilliance is determined by machine vision algorithm, and in the highest region of lamps and lanterns image brilliance
Determine that one piece is greater than or equal to the square area of 50*50 pixel as region-of-interest (RoI), to the single-pass in region-of-interest
Road/triple channel image data carries out access matrix norm calculation, its average brightness/each channel average color can be characterized by respectively obtaining
The norm of degree.
Step 4: establishing and training lamps and lanterns luminous flux degeneration/color coordinate drift model
Brightness/chroma norm based on tested lamps and lanterns in different moments can establish its luminous flux degradation model and chromaticity coordinates
Drift model.To obtain the remaining life and reliability information of lamps and lanterns, need using maximal possibility estimation, least-squares estimation,
Particle filter scheduling algorithm extracts model parameter.Such as to the luminous flux that dullness is degenerated, Gamma random process can be used establishes it and move back
Change model can be used Wiener random process and establishes its degradation model to the chromaticity coordinates of non-monotonic degeneration.
Step 5: tested lamps and lanterns predicting residual useful life and reliability assessment are carried out
Its life deterioration threshold value is determined according to the application scenarios of tested lamps and lanterns and specific requirement, the model obtained using training
Parameter and the life deterioration threshold calculations determined obtain lamps and lanterns remaining life (RUL), (tired with PDF (probability density function) and CDF
Product distribution function) characterization lamps and lanterns reliability information.
Detailed description of the invention
Fig. 1 is the overall architecture block diagram of heretofore described method
Fig. 2 is image capture device specified in the present invention and lamps and lanterns relative position explanatory diagram
Fig. 3 is image capture device specified in the present invention and lamps and lanterns relative position example figure
Fig. 4 is that optical screen schematic diagram is added specified in the present invention
Fig. 5 is qualification optical screen specified in the present invention and its region-of-interest image processing effect comparison diagram of unqualified optical screen
Fig. 6 is the machine vision part flow diagram of heretofore described method
Fig. 7 is heretofore described method image preprocessing and region-of-interest recognition effect schematic diagram
Fig. 8 is the comparison diagram of heretofore described exemplary brightness recognition result and luminous flux measurement result
Specific embodiment
To better understand technical solution of the present invention, feature and advantage, below in conjunction with attached drawing, make specifically
It is bright.
The present invention gives a kind of on-line testing methods of LED illumination lamp predicting residual useful life and reliability assessment, are
The real-time remaining life of Accurate Prediction LED illumination lamp and assessment its reliability offer support.Overall architecture of the invention is such as schemed
Shown in 1, substantive content of the invention is further illustrated with example below, but the contents of the present invention are not limited to this.
Step 1: the arrangement of image information collecting equipment and optical screen
The performance parameter for considering illuminator different mounting means and usage scenario and different images acquisition equipment, needs
Placement of images acquires equipment with wanting adaptation to local conditions, as such as solid using the image capture device met the requirements existing in scene as possible
Determine monitoring camera, fixed camera etc..The principle of image capture device arrangement is to guarantee that it can collect lamps and lanterns and be incident upon
More complete image on optical screen.If the photosensitive element height of image capture device is C, focal length f, equipment camera lens and tested
The horizontal distance of lamps and lanterns is D, and the actual height of acquired image is H, and the actual height of region-of-interest is h, then adopts to image
Collecting constraint condition when equipment is arranged isAs shown in Figure 2.The position of image capture device and mirror
Brilliance degree does not allow to change after determination, otherwise will lead to prediction result misalignment.Meanwhile image capture device need to have stabilization
Lasting power supply supply, to guarantee that it can be via manual operation or the static state of the tested lamps and lanterns of automatic collection Fixed Time Interval
Image information, acquired image information resolution need to be greater than 72ppi, and include R (red), G (green), B (blue) three
Color Channel.Optical screen is that the nearest position of the tested lamps and lanterns of distance can be connected to the smooth of the more complete picture of lamps and lanterns with what angle was fixed
Plane.According to the actual situation, optical screen can be indoor decoration plane such as metope, ground, ceiling etc., and fixed article plane is for example big
Type furniture shell, large electric appliances outer wall etc. or other planes for being used to acquire lamps and lanterns image artificially arranged.Optical screen surface it is anti-
Light characteristic needs consistent, it may be assumed that surfacing cannot have apparent concave-convex or fold, color that cannot uniformly have obvious discoloration.It is typical
Using indoor wall as optical screen image capture device arrangement example see Fig. 3.If not having in scene may be used as optical screen
Plane, fixed reflective surface can be installed additional on lamps and lanterns as additional optical screen, the square wide with lamp luminescence part is smooth
Plane is fixed on lamp installation base, guarantees that it is at least contour with lamp luminescence center, and is in 90 ° to 135 ° with lamps and lanterns horizontal line
Angle, to guarantee that the image of lamps and lanterns can completely express the light output information of lamps and lanterns.The loading pattern of additional optical screen is shown in figure
4。
Example 1: tested lamps and lanterns are a common illumination downlight on the market, and nominal information is shown in Table 1.The image of use is adopted
Integrate equipment as Canon EOS 70D digital single-lens reflex camera, Canon EF 24-105mm f/4L standard lens is installed.Its photosensitive member
Part height is 15mm, and lens focus is set as 28mm, and the linear distance with tested lamps and lanterns is 1.43m.Lamps and lanterns are tested using controlling
The mode of operating current simulates light flux variations in actual use.It is acquired using image capture device in different electric currents
The still image of the tested lamps and lanterns of lower work.
Table 1 is tested lamps and lanterns essential information
| Attribute | Parameter | Attribute | Parameter |
| Nominal power | 3.5W | Light source type | LED |
| Nominal colour temperature | 6500K | Lampshade material | PVC |
| Nominal luminous flux | 410lm | Lamp cap diameter | 110mm |
Step 2: image information pretreatment
In actual use, due to being tested the shake of lamps and lanterns, image capture device, the reasons such as scene light variation may be made
It is had a certain difference at position, the brightness etc. of lamps and lanterns image in the image of different moments acquisition.It, will based on machine vision technique
Acquire the size of still image, contrast, brightness carry out unified adjust and guarantee that its most dark areas brightness is 0 brightest area brightness
It is 255.And be sharpened, be filtered, guarantee lamps and lanterns image contour edge clear；According to the position for the lamps and lanterns image that identification obtains
Set up vertical datum mark, the image of alignment different moments acquisition.After the completion of above step, treated single channel (ash is generated
Degree) object that is identified as luminous flux information of picture, and retain the object that triple channel picture is identified as chromaticity coordinates.If optical screen matter
Amount is not able to satisfy requirement specified in this patent, such as there is protrusion, fold, then cannot may still obtain after the step process
Qualified image, it is qualified as shown in Figure 5 with region-of-interest recognition effect comparison that is not conforming to table images.
Example 2 connects example 1.Machine vision is executed using the open sources packet such as OpenCV, numpy, skimage, pandas in Python
Relevant operation.The tested lamps and lanterns still image collected under different electric currents is uniformly proceeded as follows: guaranteeing length-width ratio
Size reduction is the 50% of original image simultaneously；By 1.5 times that setting contrast is original image, brightness reduces by 1.5 units；Using corrosion
The mode of expansion filtering reduces the edge noise of image, and noise reduction rank is 1.Identify the axis of symmetry, that is, lamps and lanterns image in image
Central axis abscissa is as datum mark abscissa, and the ordinate of most bright spot is as datum mark ordinate in the axis of symmetry, with base
It is aligned the image of tested lamps and lanterns different moments on schedule.The picture that processing is completed finally is converted into single channel (i.e. luminance channel)
Grayscale image is separately deposited.
Step 3: brightness/chroma information identification
Due to the picture that self luminous lamps and lanterns itself existing in still image collected also have it to reflect on optical screen, wherein
Brightness of the lamps and lanterns in grayscale image itself is maximum 255, does not have variation tolerance, therefore do not have identification value；And lamps and lanterns
The brightness of image from closely to far gradually decreasing, therefore needs in still image collected according to its distance apart from lamps and lanterns itself
The position of lamps and lanterns image determines one piece of fixed region with brightness change tolerance as region-of-interest (RoI), the region
Maximum brightness cannot reach maximum 255, minimum brightness cannot also reach minimum 0, while need comprising enough pixels
Point is to be identified.Therefore the square area of selected lamps and lanterns image brightest area center 50*50 pixel is as region-of-interest.It will
The single channel brightness/chroma information for 2500 pixels for including in region-of-interest is considered as the matrix of a 50*50, calculates separately
Its Infinite NormIt obtains that its average brightness/each channel average chrominance norm can be characterized.Machine in the present invention
The process of device visual component, that is, step 3 and step 4 is as shown in Figure 6.
Example 3 connects example 2.In the position by finding lamps and lanterns and lamps and lanterns image in pretreated still image, and in lamps and lanterns
The most bright regional center of image determines that a block size is the region-of-interest of 50*50 pixel.The lamps and lanterns are calculated separately in different electric currents
Under the conditions of still image collected region-of-interest brightness norm, calculated result is shown in Table 2.Image preprocessing result and concern
Region is as shown in Figure 7.
Brightness norm recognition result under the different brightness of table 2
| Electric current (mA) | Norm | Electric current (mA) | Norm |
| 0 | 0 | 125 | 1882.75 |
| 50 | 1022.75 | 150 | 2007.25 |
| 75 | 1423.00 | 175 | 2053.25 |
| 100 | 1771.75 | 200 | 1966.00 |
Step 4: establishing and training lamps and lanterns luminous flux degeneration/color coordinate drift model
Brightness/chroma norm and lamps and lanterns luminous flux/chromaticity coordinates parameter that step 3 obtains are subjected to mapping matching.Compared to
Luminous flux/chromaticity coordinates absolute figure, the present invention are concerned with its opposite variation.Using the brightness/chroma initially measured as just
Initial value X0, the data X that later measures different momentsiIt is uniformed with initial value, obtains canonical measure value laterWhereinEstablish the brightness/chroma degradation model of tested lamps and lanterns respectively using canonical measure value, test proves that, the mould
Type can relatively accurately reflect the variation of tested lamps and lanterns luminous flux/chromaticity coordinates.In order to tested lamps and lanterns service life and reliability
It is predicted and is assessed, need to carry out it based on random process service life and Reliability modeling, and using measurement data to model
It is trained.
Example 4 connects example 3.The brightness norm measured is standardized first, standardization result is shown in Table 3.Observation caliberization it
It can be found that as electric current increases, the standardization norm of tested lamps and lanterns gradually increases test data afterwards, reaches in electric current
It mutates after 200mA.It is gradually degenerated to the test data simulation lamps and lanterns under 175~50mA electric current from serviceable condition luminous flux
50% process of original state.Its service life and reliability model are established using Gamma random process, if different moments are tested lamp
Tool luminous flux amount of degradation is that X (t) is the stochastic variable for obeying Gamma distribution, then leads in the light that different moments are tested lamps and lanterns
Measuring amount of degradation is the Gamma process with form parameter α and scale parameter β.The model is general different moments amount of degradation
Rate distribution function are as follows:
Wherein
Its probability density function are as follows:
The model is respectively as follows: in the expectation of t moment and variance
E [X (t)]=β α t, Var [X (t)]=β2·αt。
Its mean time to failure (MTTF) i.e. calculation formula of remaining life (RUL) are as follows:
Wherein ρ is the failure threshold of amount of degradation.
Training process to the model is to be made using the process of measurement data estimation the unknown parameter α and β at multiple moment
Measurement data is more, and model training result is better.Due to utilizing curent change to simulate degenerative process in this example, therefore cannot be right
The model is trained.But it is available with the actual luminous flux (downlight test) that laboratory installation measures tested lamps and lanterns
The brightness norm (downlight identify) obtained with identification compares, to verify the model in this example.Verifying knot
Fruit is as shown in Figure 8.
Brightness norm recognition result under the different brightness of table 3
| Electric current (mA) | Standardize norm | Electric current (mA) | Standardize norm |
| 0 | 0 | 125 | 0.92 |
| 50 | 0.50 | 150 | 0.98 |
| 75 | 0.69 | 175 | 1 |
| 100 | 0.86 | 200 | 0.96 |
Step 5: tested lamps and lanterns predicting residual useful life and reliability assessment are carried out
The remaining life and reliability of tested lamps and lanterns are predicted respectively with reliability model using the service life established
And assessment.Luminous flux or chromaticity coordinates are chosen as its parameter degenerated is measured, at certain generally according to the application scenarios of tested lamps and lanterns
A little application scenarios such as exhibition centers may also need to consider simultaneously two kinds of degradation parameters.Lamps and lanterns are tested for different degradation parameters
Lifetime threshold is different, and the hit time of lifetime threshold is its remaining life.For needing to consider the scene of two kinds of degradation parameters
Then need to find most short hit time as its remaining life.It is invoked at step respectively according to selected degradation parameter and lifetime threshold
The model established in four is brought into and obtained model parameter and threshold calculations is trained to obtain lamps and lanterns remaining life information, (general with PDF
Rate density function) with CDF (cumulative distribution function) characterization lamps and lanterns reliability information.
Example 5 connects example 4.The service life constructed using Gamma process and reliability model, bring into lifetime threshold carry out the service life with
Reliability prediction.Its two classes CDF is degradation ratio function FT(t) and Reliability Function R (t) is as follows:
Wherein ρ is lifetime threshold；R (t)=1-FT(t)。
FT(t) PDF are as follows:
The function can be used to characterize tested lamps and lanterns in the reliability of different moments.
Claims (6)
1. LED illumination lamp predicting residual useful life and reliability estimation method based on machine vision, it is characterised in that: it includes
Following steps:
Step 1: the arrangement of image information collecting equipment: according to complicated illuminator application scenarios, if according to our law regulation
Standard arrangement image capture device, image capture device need can acquisition resolution greater than 72ppi RGB triple channel image believe
Breath.
Step 2: image information pre-processes: to the image information collected, successively carrying out image scaled, contrast/bright
Degree adjustment, the determination of identification datum mark and gray proces, obtain the qualified image information of processing.
Step 3: brightness/chroma information identifies: reflecting lamp brightness/coloration information in identification image information, determine concern area
Domain carries out matrix operation to the data in region-of-interest and obtains its norm.
Step 4: establish and train lamps and lanterns luminous flux degenerate/coordinate shift model: according to the brightness/chroma model of identification arrived
Number is established the lamps and lanterns luminous flux/chromaticity coordinates degradation model characterized by it respectively using random process, and is based on maximum likelihood
Model parameter is extracted in the methods of estimation, least square method, particle filter.
Step 5: carrying out tested lamps and lanterns predicting residual useful life and reliability assessment: determining that luminous flux moves back respectively for different lamps
Change/color coordinate drift threshold value utilizes the model parameter further progress lamps and lanterns predicting residual useful life and reliability assessment extracted.
2. a kind of illuminator predicting residual useful life and reliability assessment side based on machine vision according to claim 1
Method, it is characterised in that: in the first step in " arrangement of image information collecting equipment ", consider the peace of illuminator complexity
Dress mode and usage scenario utilize the image capture device such as monitoring camera, digital camera etc. that position, shooting angle are fixed
Complete the acquisition of image information；Image capture device directly shoots lamps and lanterns and lamps and lanterns project the shadow on nearest fixation optical screen
Picture；Image capture device is acquired the lamps and lanterns static image information of Fixed Time Interval and is stored up by way of manual or automatic control
It deposits.
3. a kind of illuminator predicting residual useful life and reliability assessment side based on machine vision according to claim 1
Method, it is characterised in that: in " the image information pretreatment " described in second step, the image collected by the first step is believed
Breath scaling reaches unified size；According to the average brightness difference of brightest area in image and most dark areas, it is whole to carry out image
Contrast and brightness adjustment；It identifies the lamps and lanterns center position in image, is acquired as benchmark point alignment different time
Obtained image；The object that the corresponding grayscale image of image is identified as luminous flux information is generated using channel transfer algorithm.
4. a kind of illuminator predicting residual useful life and reliability assessment side based on machine vision according to claim 1
Method, it is characterised in that: in the third step in " identification of brightness/chroma information ", using machine vision algorithm, to passing through the
The image data that two steps are handled is identified that one piece for obtaining being located at lamps and lanterns image brilliance highest zone is greater than or equal to
The square area of 50*50 pixel size is as region-of-interest；Access matrix norm is carried out to the image data in region-of-interest
It calculates, its average brightness/each channel average chrominance norm can be characterized by respectively obtaining.
5. a kind of illuminator predicting residual useful life and reliability assessment side based on machine vision according to claim 1
Method, it is characterised in that: in " establishing described in the 4th step and train lamps and lanterns luminous flux degeneration/color coordinate drift model ", lead to
It crosses the brightness/chroma norm that third step identifies and establishes characterization luminous flux/chromaticity coordinates degradation model；It is random based on Gamma
Process establishes lamps and lanterns luminous flux degeneration service life and reliability model, establishes the lamps and lanterns color coordinate drift longevity based on Wiener random process
Life and reliability model, and it is utilized respectively the method training pattern parameter of Maximum-likelihood estimation.
6. a kind of illuminator predicting residual useful life and reliability assessment side based on machine vision according to claim 1
Method, it is characterised in that: in described in the 5th step " progress predicting residual useful life and reliability assessment ", not according to illuminator
Same application scenarios are judged as luminous flux sensitivity scene or chromaticity coordinates sensitivity scene or comprehensive sensitive scene, determine lamps and lanterns respectively
Life deterioration threshold value；It predicts to obtain lamps and lanterns remaining life information using the prediction model that the 4th step obtains, with PDF, (probability is close
Spend function) with CDF (cumulative distribution function) characterization lamps and lanterns reliability information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910322099.XA CN110062502B (en) | 2019-04-22 | 2019-04-22 | Machine vision-based online residual life prediction and reliability evaluation method for LED lighting lamp |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910322099.XA CN110062502B (en) | 2019-04-22 | 2019-04-22 | Machine vision-based online residual life prediction and reliability evaluation method for LED lighting lamp |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110062502A true CN110062502A (en) | 2019-07-26 |
| CN110062502B CN110062502B (en) | 2020-05-19 |
Family
ID=67319905
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910322099.XA Active CN110062502B (en) | 2019-04-22 | 2019-04-22 | Machine vision-based online residual life prediction and reliability evaluation method for LED lighting lamp |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110062502B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110455503A (en) * | 2019-08-15 | 2019-11-15 | 亿轶环境科技(上海)有限公司 | A kind of ultraviolet tube service life monitoring method and device |
| CN110737987A (en) * | 2019-10-16 | 2020-01-31 | 北京航空航天大学 | Method for evaluating expected life of LED lighting products |
| CN111854939A (en) * | 2020-07-24 | 2020-10-30 | 深圳市明学光电股份有限公司 | Online detection method for LED flexible light bar |
| CN113673721A (en) * | 2021-08-26 | 2021-11-19 | 北京航空航天大学 | Cluster system preventive maintenance method based on deep reinforcement learning |
| CN116017829A (en) * | 2022-12-30 | 2023-04-25 | 亚翔系统集成科技（苏州）股份有限公司 | Dust-free room lamp service life monitoring method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003180632A (en) * | 2001-12-20 | 2003-07-02 | Pentax Corp | Lamp remaining life calculation device of endoscopic equipment |
| CN102974551A (en) * | 2012-11-26 | 2013-03-20 | 华南理工大学 | Machine vision-based method for detecting and sorting polycrystalline silicon solar energy |
| CN104034516A (en) * | 2014-06-23 | 2014-09-10 | 华高科技（苏州）有限公司 | Machine vision based LED detection device and detection method thereof |
| CN105241638A (en) * | 2015-09-09 | 2016-01-13 | 重庆平伟光电科技有限公司 | Vision-based quick LED module brightness uniformity detection method |
| CN105468866A (en) * | 2015-12-15 | 2016-04-06 | 长春工业大学 | Method for predicting remaining life of LED driving power of railway vehicles |
| US20160219669A1 (en) * | 2014-03-10 | 2016-07-28 | Dynotron, Inc. | Variable lumen output and color spectrum for led lighting |
| CN108135052A (en) * | 2017-12-20 | 2018-06-08 | 中国电子产品可靠性与环境试验研究所 | The status information monitoring method of LED lamp, apparatus and system |
-
2019
- 2019-04-22 CN CN201910322099.XA patent/CN110062502B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2003180632A (en) * | 2001-12-20 | 2003-07-02 | Pentax Corp | Lamp remaining life calculation device of endoscopic equipment |
| CN102974551A (en) * | 2012-11-26 | 2013-03-20 | 华南理工大学 | Machine vision-based method for detecting and sorting polycrystalline silicon solar energy |
| US20160219669A1 (en) * | 2014-03-10 | 2016-07-28 | Dynotron, Inc. | Variable lumen output and color spectrum for led lighting |
| CN104034516A (en) * | 2014-06-23 | 2014-09-10 | 华高科技（苏州）有限公司 | Machine vision based LED detection device and detection method thereof |
| CN105241638A (en) * | 2015-09-09 | 2016-01-13 | 重庆平伟光电科技有限公司 | Vision-based quick LED module brightness uniformity detection method |
| CN105468866A (en) * | 2015-12-15 | 2016-04-06 | 长春工业大学 | Method for predicting remaining life of LED driving power of railway vehicles |
| CN108135052A (en) * | 2017-12-20 | 2018-06-08 | 中国电子产品可靠性与环境试验研究所 | The status information monitoring method of LED lamp, apparatus and system |
Non-Patent Citations (1)
| Title |
|---|
| CHENG QIAN ET AL.: "Photometric and Colorimetric Assessment of LED Chip Scale Packages by Using a Step-Stress Accelerated Degradation Test (SSADT) Method", 《MATERIALS》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110455503A (en) * | 2019-08-15 | 2019-11-15 | 亿轶环境科技(上海)有限公司 | A kind of ultraviolet tube service life monitoring method and device |
| CN110455503B (en) * | 2019-08-15 | 2021-09-17 | 亿轶环境科技(上海)有限公司 | Ultraviolet tube service life monitoring method and device |
| CN110737987A (en) * | 2019-10-16 | 2020-01-31 | 北京航空航天大学 | Method for evaluating expected life of LED lighting products |
| CN111854939A (en) * | 2020-07-24 | 2020-10-30 | 深圳市明学光电股份有限公司 | Online detection method for LED flexible light bar |
| CN111854939B (en) * | 2020-07-24 | 2023-02-03 | 深圳市明学光电股份有限公司 | Online detection method for LED flexible light bar |
| CN113673721A (en) * | 2021-08-26 | 2021-11-19 | 北京航空航天大学 | Cluster system preventive maintenance method based on deep reinforcement learning |
| CN116017829A (en) * | 2022-12-30 | 2023-04-25 | 亚翔系统集成科技（苏州）股份有限公司 | Dust-free room lamp service life monitoring method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110062502B (en) | 2020-05-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110062502A (en) | A kind of online predicting residual useful life of LED illumination lamp based on machine vision and reliability estimation method | |
| CN1322307C (en) | Apparatus for surface inspection and method and apparatus for inspecting substrate | |
| CN105973571B (en) | A kind of measuring method of the LED chip macroscopic surface brightness based on CCD | |
| CN106093073B (en) | Base board defect location positioning method and device and system | |
| CN106104375B (en) | The flashing light light of tool optimization wave spectral power distributions | |
| CN107543607A (en) | A kind of method, system and equipment for detecting lighting environment health indicator | |
| CN109387281A (en) | The luminosity with Real-time Feedback monitors system in luminous environment simulation | |
| CN104316295A (en) | Photoelectric test method and device of LED device | |
| CN114136975A (en) | Intelligent detection system and method for surface defects of microwave bare chip | |
| Ismail et al. | Development of a webcam based lux meter | |
| CN111261079A (en) | Detection method for abnormal phenomena of bright spots and dark spots | |
| US20130262006A1 (en) | White LED Quality Inspection Method and Device | |
| CN114867161A (en) | Commercial lighting system and lighting method based on AIoT and sensor network | |
| CN110831276A (en) | LED-based lamplight brightness control method and related device | |
| CN109803465A (en) | Illumination and color temperature compensating control method and system based on room lighting | |
| CN108696960A (en) | The adjusting method and roadway lighting system of roadway lighting system | |
| CN114022820A (en) | Intelligent beacon light quality detection method based on machine vision | |
| CN111721507A (en) | Intelligent detection method and device for keyboard backlight module based on polar coordinate identification | |
| CN217542344U (en) | AR lens transmission performance test equipment | |
| CN110216082A (en) | The recognition methods of fluorescent marker seed dynamics and system | |
| CN106033062A (en) | Automatic dimming method for optical detection and optical detection machine platform thereof | |
| CN109708752A (en) | A kind of method of feedback compensation in skylight environmental simulation | |
| Zatari et al. | Glare, luminance, and illuminance measurements of road lighting using vehicle mounted CCD cameras | |
| Rykowski et al. | Novel approach for LED luminous intensity measurement | |
| CN209014481U (en) | Vision detection system |
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 |