CN110232660A - A kind of new infrared image recognition pretreatment gray scale stretching method - Google Patents
A kind of new infrared image recognition pretreatment gray scale stretching method Download PDFInfo
- Publication number
- CN110232660A CN110232660A CN201910347677.5A CN201910347677A CN110232660A CN 110232660 A CN110232660 A CN 110232660A CN 201910347677 A CN201910347677 A CN 201910347677A CN 110232660 A CN110232660 A CN 110232660A
- Authority
- CN
- China
- Prior art keywords
- image
- gray scale
- temperature
- value
- processing
- 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
- 238000000034 method Methods 0.000 title claims abstract description 15
- 238000012545 processing Methods 0.000 claims abstract description 50
- 238000001514 detection method Methods 0.000 abstract description 6
- 238000012549 training Methods 0.000 description 6
- 238000013135 deep learning Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000007613 environmental effect Effects 0.000 description 4
- 238000003384 imaging method Methods 0.000 description 4
- 238000007781 pre-processing Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 239000002956 ash Substances 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000003672 processing method Methods 0.000 description 2
- 230000000306 recurrent effect Effects 0.000 description 2
- 238000001931 thermography Methods 0.000 description 2
- WJCNZQLZVWNLKY-UHFFFAOYSA-N thiabendazole Chemical compound S1C=NC(C=2NC3=CC=CC=C3N=2)=C1 WJCNZQLZVWNLKY-UHFFFAOYSA-N 0.000 description 2
- 241001270131 Agaricus moelleri Species 0.000 description 1
- 235000002918 Fraxinus excelsior Nutrition 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000007667 floating Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007935 neutral effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of new infrared image recognitions to pre-process gray scale stretching method, belongs to technical field of image processing.The present invention can be used for the detection of the constant temperature object in infrared image, such as pedestrian's identification, animal identification, vehicle identification etc..The present invention first corresponds temperature and gray scale, the gray scale details of object is improved by being stretched gray scale within the temperature range of identified object, the gray feature for reducing identified object is influenced by environment temperature, to improve the universality for carrying out gray scale stretching processing for image, and need to carry out image grayscale in the object identification system in promotion infrared image and stretch pretreated processing links, to promote the recognition accuracy to target object.
Description
Technical field
The present invention relates to technical field of image processing, and in particular to the gray scale stretching processing technique of a kind of pair of infrared image.
Background technique
The popular domain of pedestrian's identification always image recognition in infrared image, and the field that has a wide range of applications, such as
What improves the accuracy of identification of infrared image, and recognition accuracy is always important research problem in industry.
In pedestrian's identifying processing of infrared image, acquisition infrared image and image preprocessing (including gray scale is carried out first
Stretch processing), it is then based on preset identification model (usually deep learning network), corresponding training dataset (instruction is set
Practice sample standard deviation and pass through the image preprocessing comprising stretch processing), and identification model is trained based on it；Finally, utilizing instruction
Identification model is perfected, image to be detected after the image preprocessing handled comprising gray scale stretching is inputted into trained identification
Model carries out the processing of pedestrian's recognition detection.
Wherein, existing image grayscale stretches the specific processing mode of (also referred to as image grayscale is shown) scheme are as follows:Wherein, G is the gray value of respective coordinates point in image, and TfmaxTo be passed in the image
Maximum temperature in the image of sensor identification, TfminThe minimum temperature in image is identified for sensor in the image, and t is coordinate points
Sensor receive temperature value.
The main thought of i.e. existing image grayscale stretch processing is that maximum temperature and minimum temperature are divided into 256 ashes
Then the gray value of temperature value and picture that sum of the grayscale values sensor identifies is mapped to constant gradient, reaches imaging by angle value
Purpose.But the processing mode has larger defect in object identification, the identification when excessive temperature differentials, for cryogenic object
It is more difficult, it is difficult to be separated, when environment temperature and excessively close identified temperature, also will affect to identified object
Detection.I.e. it lacks the different situations according to image and carries out self-control and to reach preferably showing to object to be detected,
It is flexibly poor.
Since the image grayscale stretch processing scheme in the current processing scheme for the identification of pedestrian lacks for one
The universality of the various situations in year four seasons, so that it is larger or when environment temperature is close easily identifies not in environmental difference
The technical issues of being identified object or wrong identification out.
Summary of the invention
Goal of the invention of the invention is: in view of the above problems, providing a kind of new infrared image recognition and locates in advance
Manage gray scale stretching method.
The technical solution adopted by the present invention are as follows:
The temperature of sensor based on image to be processed (image of pending gray scale stretching processing) receives range [Tmin,
Tmax], the gray value R that each pixel of image to be processed is shown on the image is set are as follows:
Wherein t is the temperature value for each pixel that sensor receives；
Stretch processing is carried out to gray value R, obtains gray value g, whereinThe value of parameter m, E are respectively as follows:
The median setting parameter m for showing temperature range based on constant temperature object to be identified；
According to preset gray scale stretching range [K1,K2] and parameter m setting parameter E value:
Calculating parameter
According to formula E=min { E1, E2Calculating parameter E value；
The final gray value G of each pixel that image to be processed is obtained based on G=255 × g, obtains the ash of image to be processed
Spend stretch processing image.
In conclusion by adopting the above-described technical solution, the beneficial effects of the present invention are:
The present invention combines on the basis of existing gray scale stretching processing scheme, by it with the pixel temperatures of image pair
The temperature range of object to be identified is stretched, and improves details and the bright-dark degree of identified object, and environment gap compared with
There is good capacity of self-regulation when big.To effectively improve in the target identification processing scheme in infrared image in ring
Border temperature difference is excessive or identified object and object identification accuracy rate when close environment temperature.
Detailed description of the invention
Fig. 1 is existing gray scale stretching processing method treated image effect schematic diagram.
Fig. 2 is gray scale stretching processing method of the invention treated image effect schematic diagram.
Fig. 3 is the recognition result schematic diagram to gray scale stretching shown in FIG. 1 processing image.
Fig. 4 is the recognition result schematic diagram to gray scale stretching shown in Fig. 2 processing image, wherein identifying processing mode and figure
3 is identical.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below with reference to embodiment and attached drawing, to this hair
It is bright to be described in further detail.
The present invention can be used for the detection of the constant temperature object in infrared image, such as pedestrian's identification, animal identification, vehicle identification etc.
Aspect, based on the pretreated constant temperature object identification processing of gray scale stretching of the invention, (identifying processing scheme can be used existing
Target identification processing scheme in infrared image, but the gray scale stretching processing in image preprocessing uses gray scale stretching of the invention
Processing mode), condition corresponding target object can be also accurately identified similar in or environment temperature larger in environmental difference.
In order to realize gray scale stretching processing of the invention, temperature and gray scale are corresponded first, by that will be known
Within the temperature range of other object gray scale stretched reduced to improve the gray scale details of object the gray feature of identified object by
The influence of environment temperature to improve the universality for carrying out gray scale stretching processing for image, and is promoted in infrared image
It needs to carry out image grayscale in object identification system and stretches pretreated processing links, so that the identification promoted to target object is quasi-
Exactness.
In the identifying system that Digital Image Processing and artificial intelligence combine, in order to realize in infrared hybrid optical system
The detection (such as pedestrian) of constant temperature object, it is contemplated that constant temperature object is relatively stable radiation source, and surface temperature is in a Nian Si
Season, there is no big floatings, were influenced also not being very big by environment temperature.So sensor is received temperature model by the present invention
The intensity value ranges for enclosing and showing on the image establish one-to-one relationship, so that the image that constant temperature object is imaged in sensor
On imaging effect it is relatively stable.
Sensor-based temperature receives range [Tmin,Tmax], the gray value R that each pixel is shown on the image is set
Are as follows:Wherein t is the temperature value that sensor receives the pixel in region.So that
The temperature that each sensor receives corresponds to a specific gray value.
It but is only far from being enough, deficiency by the gray value one-to-one correspondence of the temperature range of sensor measurement and image
Place shows all gray scales that image can not be made full use of to show, because -30 degrees Celsius and 350 Celsius in natural environment
These extreme temperatures are spent so uncommon, so the present invention is by the corresponding gray scale of radiometric temperature range of constant temperature object
It stretches with carrying out specific shape, it is made to use image explicitly range as far as possible, in the hope of showing the specific thin of more constant temperature objects
Section, to improve the accuracy of subsequent identification.
Gray value after stretching indicates with g, specifically:
The range of gray value R is stretched, stretches the curve that image is s shape.Wherein, m, E determine stretching
The shape of curve, value are derived by its numerical value according to the temperature range of identified object.
Show temperature range [T the present invention is based on constant temperature object to be identifiedo_minTo_max], the value of parameter m is set
Are as follows: the median for showing temperature range of constant temperature object to be identified, i.e. To_minAnd To_maxMedian (intermediate value)；And it is based on
Preset gray scale stretching range [K1,K2] and parameter m setting parameter E value are as follows:
E=min { E1, E2}；
Finally, gray scale stretching of the invention is obtained treated gray value G based on G=255 × g, i.e., it will be in [0,1] range
Interior gray value g expands to [0,255] totally 256 gray values.
That is, the present invention is in order to realize the gray scale stretching processing in infrared image identification pretreatment, it is first sensor receiving
Temperature and the gray scale of presentation establish a corresponding relationship, and the gray scale model of detection object temperature range is then found by corresponding relationship
It encloses, then by tonal range with being stretched to image entire indication range, in the hope of preferably being shown to identified object.
Embodiment
Constant temperature object to be detected is pedestrian, and it is -30 DEG C to 350 DEG C that the temperature of used sensor, which receives range, therefore
T is setmax=350, Tmin=-30；For winter, the surface temperature of pedestrian is usually between 10 DEG C to 40 DEG C, therefore taking m is 10 DEG C
With 26 DEG C of intermediate value of 40 DEG C, gray scale stretching range is set as K1=0.01, K2=0.99；Based on the set-up mode of parameter E, can obtain
To E=7.
In order to realize to pedestrian's identifying processing in infrared image, the present embodiment mainly includes three processing links；Data
Collection is established, deep learning environmental structure (identification model based on deep learning network) and target training identification.
It is the acquisition of samples pictures first, the present embodiment uses Testo875-i thermal infrared in data set foundation
Imager, the imaging temperature range of thermal imaging system are -30 DEG C to 350 DEG C, and photo resolution is 160 × 120, and collecting location is school
Garden building is other, has collected 1260 pictures altogether.
Followed by the secondary treatment to picture, since the collected picture of thermal imaging instrument is that Testo company is distinctive
Picture format, the picture export software I Rsoft that the present invention needs to borrow Testo will acquire picture export, and export result is
Excel table, the secondary treatment in order to the later period to acquisition image.
After derived form document is read out using Matlab software, be respectively adopted the processing of existing gray scale stretching and
The improved gray scale stretching processing of the present invention handles list data, and processing result difference is as illustrated in fig. 1 and 2；And turn
Imaging saves as two groups after turning to jpeg file format, and one group is existing gray scale stretching processing result, and one group is of the invention
Gray scale stretching processing result.
The foundation of data set needs for data set to be divided into training set and test set, and intensive 1000 picture of training uses mark
Label tool labelimg is visually demarcated to two groups of pictures are obtained, and calibration object is all corresponding training set in two groups, is left
260 pictures be test set, be used for test verification.Data set is using Pascal VOC format.
It then, is deep learning environmental structure, the environment of deep learning has very much, such as TensorFlow, Keras,
Caffe etc., the present embodiment is using TensorFlow frame, because the development of the frame is more rapid, maturity sum aggregate
It is neutral higher.
Followed by the selection of specific recognizer, alternative algorithm has Fast-RCNN, Yolo, Faster R-CNN
Deng, in the present embodiment, the recognizer used is Faster R-CNN, because the recognition accuracy of the algorithm is very high, Er Qieshi
Other speed is fast, has very high Practical significance, and the computer system of operation is ubuntu 18.04, and video card model is tall and handsome reaches
1050Ti。
It is finally target training identification, Faster R-CNN network includes 3 parts, including feature extraction network, RPN
(RegionProposal Network) network, Recurrent networks of classifying.
Wherein, the selection of feature extraction network can very flexibly, and the present embodiment uses VGG16 network.
The major function of RPN network is handled the output of feature extraction network, is obtained output classification, is then generated
The RoI (region of interest) and corresponding label information that next network needs.
Classification Recurrent networks, the confidence level classified after determining loss function to detected picture, and classified,
Export identified picture.
It is compared by emulation experiment, gray scale stretching processing proposed by the present invention and the processing of existing gray scale stretching exist centainly
Advantage, such as the comparison between Fig. 1 and Fig. 2 it can be found that gray scale stretching processing of the invention is for being identified the display of image more
Add clear and definite, is easy identification, is not influenced by highest temperature object in environment (most bright place is lamp).From the knot of Fig. 3 and Fig. 4
From the point of view of fruit figure comparison diagram, the identifying processing result based on gray scale stretching of the invention treated image not known originally
Not Chu Lai the pedestrian together with another pedestrian identify, achieve better recognition result.
For the more intuitive recognition performance more of the invention for target to be identified, using mAP (mean Average
Precision) the index as the precision for measuring target identification, which show the precision of target identification and accuracys.
In the present embodiment, the mesh based on identifying processing of the invention with the identifying processing based on the processing of existing gray scale stretching
It is as shown in table 1 to mark recognition performance:
Table 1
| Method | Existing method | The present invention |
| mAP | 87.71% | 88.54% |
As shown in Table 1, based on identifying processing of the invention than the identifying processing side that is handled based on existing gray scale stretching
Method has the raising of nearly one percentage point, obvious to the performance boost of target identification.
The above description is merely a specific embodiment, any feature disclosed in this specification, except non-specifically
Narration, can be replaced by other alternative features that are equivalent or have similar purpose；Disclosed all features or all sides
Method or in the process the step of, other than mutually exclusive feature and/or step, can be combined in any way.
Claims (2)
1. a kind of new infrared image recognition pre-processes gray scale stretching method, characterized in that it comprises the following steps:
The temperature of sensor based on image to be processed receives range [Tmin, Tmax], each pixel that image to be processed is arranged exists
The gray value R shown on image are as follows:Wherein t is each pixel that sensor receives
Temperature value；
Stretch processing is carried out to gray value R, obtains gray value g, whereinThe value of parameter m, E are respectively as follows:
The median setting parameter m for showing temperature range based on constant temperature object to be identified；
According to preset gray scale stretching range [K1, K2] and parameter m setting parameter E value:
Calculating parameter
According to formula E=min { E1, E2Calculating parameter E value；
The final gray value G of each pixel that image to be processed is obtained based on G=255 × g, the gray scale for obtaining image to be processed are drawn
Stretch processing image.
2. the method as described in claim 1, which is characterized in that gray scale stretching range [K1, K2] setting are as follows: K1=0.01, K2=
0.99。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910347677.5A CN110232660B (en) | 2019-04-28 | 2019-04-28 | Novel infrared image recognition preprocessing gray stretching method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910347677.5A CN110232660B (en) | 2019-04-28 | 2019-04-28 | Novel infrared image recognition preprocessing gray stretching method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110232660A true CN110232660A (en) | 2019-09-13 |
| CN110232660B CN110232660B (en) | 2022-05-03 |
Family
ID=67860302
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910347677.5A Active CN110232660B (en) | 2019-04-28 | 2019-04-28 | Novel infrared image recognition preprocessing gray stretching method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110232660B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110824311A (en) * | 2019-11-22 | 2020-02-21 | 南京信息工程大学 | Zinc oxide valve plate breakdown point detection device and method based on image recognition |
| CN111724301A (en) * | 2020-06-19 | 2020-09-29 | 电子科技大学 | Self-adaptive stretching method and system based on histogram statistics |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090033835A1 (en) * | 2005-02-25 | 2009-02-05 | Fujifilm Corporation | Optical Compensation Sheet, Polarizing Plate and Liquid Crystal Display |
| US20090174712A1 (en) * | 2006-07-31 | 2009-07-09 | Sandviken Intellectual Property Ab | Method, apparatus and computer-readable medium for scale-based visualization of an image dataset |
| CN102222323A (en) * | 2011-06-13 | 2011-10-19 | 北京理工大学 | Histogram statistic extension and gradient filtering-based method for enhancing infrared image details |
| CN102798475A (en) * | 2012-08-02 | 2012-11-28 | 电子科技大学 | Method for determining temperature of object according to standard infrared video image |
| CN105354814A (en) * | 2015-08-17 | 2016-02-24 | 电子科技大学 | Infrared image segmentation method |
| CN106570831A (en) * | 2016-10-09 | 2017-04-19 | 中国航空工业集团公司洛阳电光设备研究所 | Gray image contrast equalization enhancement method |
| CN107403410A (en) * | 2017-07-25 | 2017-11-28 | 北京华新创科信息技术有限公司 | A kind of joining method of thermal infrared images |
| CN107507200A (en) * | 2017-08-31 | 2017-12-22 | 电子科技大学 | A kind of a wide range of water area extraction method of SAR image high-precisions based on communication with detection and noise suppressed |
| CN109377469A (en) * | 2018-11-07 | 2019-02-22 | 永州市诺方舟电子科技有限公司 | A kind of processing method, system and the storage medium of thermal imaging fusion visible images |
-
2019
- 2019-04-28 CN CN201910347677.5A patent/CN110232660B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090033835A1 (en) * | 2005-02-25 | 2009-02-05 | Fujifilm Corporation | Optical Compensation Sheet, Polarizing Plate and Liquid Crystal Display |
| US20090174712A1 (en) * | 2006-07-31 | 2009-07-09 | Sandviken Intellectual Property Ab | Method, apparatus and computer-readable medium for scale-based visualization of an image dataset |
| CN102222323A (en) * | 2011-06-13 | 2011-10-19 | 北京理工大学 | Histogram statistic extension and gradient filtering-based method for enhancing infrared image details |
| CN102798475A (en) * | 2012-08-02 | 2012-11-28 | 电子科技大学 | Method for determining temperature of object according to standard infrared video image |
| CN105354814A (en) * | 2015-08-17 | 2016-02-24 | 电子科技大学 | Infrared image segmentation method |
| CN106570831A (en) * | 2016-10-09 | 2017-04-19 | 中国航空工业集团公司洛阳电光设备研究所 | Gray image contrast equalization enhancement method |
| CN107403410A (en) * | 2017-07-25 | 2017-11-28 | 北京华新创科信息技术有限公司 | A kind of joining method of thermal infrared images |
| CN107507200A (en) * | 2017-08-31 | 2017-12-22 | 电子科技大学 | A kind of a wide range of water area extraction method of SAR image high-precisions based on communication with detection and noise suppressed |
| CN109377469A (en) * | 2018-11-07 | 2019-02-22 | 永州市诺方舟电子科技有限公司 | A kind of processing method, system and the storage medium of thermal imaging fusion visible images |
Non-Patent Citations (1)
| Title |
|---|
| 葛朋 等: "一种基于引导滤波图像分层的红外图像细节增强算法", 《红外技术》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110824311A (en) * | 2019-11-22 | 2020-02-21 | 南京信息工程大学 | Zinc oxide valve plate breakdown point detection device and method based on image recognition |
| CN111724301A (en) * | 2020-06-19 | 2020-09-29 | 电子科技大学 | Self-adaptive stretching method and system based on histogram statistics |
| CN111724301B (en) * | 2020-06-19 | 2023-05-12 | 电子科技大学 | Self-adaptive stretching method and system based on histogram statistics |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110232660B (en) | 2022-05-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2023029373A1 (en) | High-precision farmland vegetation information extraction method | |
| CN110880016B (en) | Image style migration method, device, equipment and storage medium | |
| CN109977780A (en) | A kind of detection and recognition methods of the diatom based on deep learning algorithm | |
| CN104298992B (en) | A kind of adaptive scale pedestrian recognition methods again based on data-driven | |
| CN104392240A (en) | Parasite egg identification method based on multi-feature fusion | |
| CN111724355B (en) | Image measuring method for abalone body type parameters | |
| Sun et al. | Wheat head counting in the wild by an augmented feature pyramid networks-based convolutional neural network | |
| CN109508709B (en) | Single pointer instrument reading method based on machine vision | |
| CN111724358A (en) | Concrete quality detection method and system based on image and convolutional neural network | |
| CN111881851A (en) | Garden seedling intelligent detection and counting method based on UAV and convolutional neural network | |
| CN110232660A (en) | A kind of new infrared image recognition pretreatment gray scale stretching method | |
| CN106803087A (en) | A kind of car number automatic identification method and system | |
| Wang et al. | Object-oriented change detection approach for high-resolution remote sensing images based on multiscale fusion | |
| CN108230307A (en) | A kind of corn broken kernel detection method based on profile centroid distance and neural network | |
| CN111882664A (en) | Multi-window accumulated difference crack extraction method | |
| CN115880594A (en) | Intelligent dam crack detection method based on unmanned aerial vehicle visual perception and deep learning | |
| CN107563327B (en) | Pedestrian re-identification method and system based on self-walking feedback | |
| CN113837166A (en) | Pointer type instrument automatic reading method based on deep learning | |
| CN112213244A (en) | Motor vehicle exhaust ringer-Mannheim blackness measuring device and method based on machine learning | |
| CN110930393A (en) | Chip material pipe counting method, device and system based on machine vision | |
| CN113688819B (en) | Target object expected point tracking and matching method based on marked points | |
| CN113627480B (en) | Polarization SAR image classification method based on reinforcement learning | |
| Wang et al. | Study on estimation method of plant leaf area based on image processing technology | |
| CN114463628A (en) | Deep learning remote sensing image ship target identification method based on threshold value constraint | |
| Yao et al. | An adaptive and quantitative rubber nutrient status analyzing system by digital foliar images |
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 |