CN106952255A - Optical remote satellite image floor treatment lifting test system - Google Patents
Optical remote satellite image floor treatment lifting test system Download PDFInfo
- Publication number
- CN106952255A CN106952255A CN201710168902.XA CN201710168902A CN106952255A CN 106952255 A CN106952255 A CN 106952255A CN 201710168902 A CN201710168902 A CN 201710168902A CN 106952255 A CN106952255 A CN 106952255A
- Authority
- CN
- China
- Prior art keywords
- image
- work station
- processing
- lifting
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/20—Processor architectures; Processor configuration, e.g. pipelining
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Processing (AREA)
Abstract
A kind of Optical remote satellite image floor treatment lifting test system, belongs to remote sensing image processing, A+E technical field.The pilot system includes data and sends work station, digital image procossing lifting work station, image procossing lifting processor, processing improving performance evaluation work station and Ethernet switch；The Ethernet switch is used to provide the Ethernet connection between data transmission work station, digital image procossing lifting work station, image procossing lifting processor, processing improving performance evaluation work station.It it is an advantage of the invention that system is complete and comprehensive, can comprehensively reflect the picture quality integrated treatment hoisting power of image processing algorithm software and engineering prototype hardware system, and more fully and effectively instruct the design, checking and evaluation work of Processing Algorithm；Hardware platform is connected by high-speed PCI E buses with each process plate, and parallel view data in equipment, in real time flowing water, processing can be achieved；Scalability is strong, ensure that the Efficient Operation of system.
Description
Technical field
The invention belongs to remote sensing image processing, A+E technical field, and in particular to a kind of optical remote sensing is defended
Star chart is as floor treatment lifting test system.
Background technology
For optical remote sensing imaging system, the usual method elimination camera optical system handled using image restoration,
Image blurring, the noise iseikonia matter caused by factor such as detector, electronics for imaging and satellite platform environment and air path is moved back
Change, and practical engineering application significant to the effectiveness that gives full play to image is worth.Foreign countries are by the image on ground
The organic and important composition link as the full link imaging system design of remote sensing is handled, restoration disposal is fully combined and carries out main chain
The matching optimization and design of road link, realize camera design transmission function it is relatively low in the case of, using subsequent treatment ensure most
Whole picture quality, the transmission function (MTF after recovery at image nyquist frequencyN) value raising more than 70%, make system development
Difficulty and cost are substantially reduced.In terms of existing literature, the research of image processing algorithm is a lot, in order at utmost in algorithm research
The process performance of boosting algorithm, generally combines the thoughts such as node-by-node algorithm, successive ignition and sets up processing model and method, this is just inevitable
Cause algorithm sufficiently complex, amount of calculation is huge when particularly handling large-sized high-resolution optical remote sensing satellite image, therefore
It is general using graphic process unit (GPU), field programmable gate array (FPGA) and multiple data signals in practical implementation
The hardware platforms such as processor (DSP), engineering is carried out to algorithm and is realized with accelerating, and develops corresponding image procossing principle sample
Machine.
Therefore, be badly in need of building it is a set of sent comprising data, digital image procossing lifting work station, at image procossing lifting
Reason machine, the ground image processing lifting test system at processing improving performance evaluation work station, treatability of the system for algorithm
It can assess, algorithm optimization design and its practical engineering application have important theory directive significance and engineering application value.
The content of the invention
The invention aims to overcome the above-mentioned problems in the prior art, there is provided a set of Optical remote satellite figure
As floor treatment lifting test system, user can select digital image procossing to be lifted at work station or image by management and control parameter
Reason lifting processor operation image processing algorithm.Wherein digital image procossing lifting work station uses Processing Algorithm software platform
Digital mode of operation handled, image procossing lifting processor is using the half of algorithm engineering accelerating hardware processing platform
Physics Work pattern is handled.Optical remote satellite image processing algorithm software, processor function can be realized using the system
With performance evaluation, evaluation and the test and validation of processor key modules, and the application boundary and optimizing research of algorithm are supported.
To achieve the above object, the technical scheme that the present invention takes is as follows：
A kind of Optical remote satellite image floor treatment lifting test system, the pilot system include data send work station,
Digital image procossing lifting work station, image procossing lifting processor, processing improving performance evaluation work station and Ethernet are handed over
Change planes；The Ethernet switch is used to provide data transmission work station, digital image procossing lifting work station, image procossing
Lift the Ethernet connection between processor, processing improving performance evaluation work station.
The present invention is relative to the beneficial effect of prior art：
（1）The present invention for the process performance analysis of image processing algorithm and its Project Realization algorithm, key modules performance test,
Checking and the demand assessed, have built a set of Optical remote satellite image floor treatment lifting test system, the system is contained
The function such as image and assistance data transmission, image characteristics extraction, image procossing and Processing Algorithm Performance Evaluation, and figure can be supported
The digital algorithm software processing of picture and the processor hardware of semi physical handle two kinds of mode of operations, possess data storage and hair
Give, image processing algorithm software and processor hardware system performance analysis and evaluation and the performance test of processor key modules with
The functions such as checking, and the application boundary and optimizing research of Processing Algorithm can be supported.
（2）Optical remote satellite image floor treatment lifting test system of the present invention includes processing improving performance evaluation work
Stand, possess image transmission function handles improving performance, the processing improving performance of signal noise ratio (snr) of image, information fidelity performance, pseudomorphism
The evaluation capacities such as rejection, evaluation index system is complete and comprehensive, can comprehensively reflect image processing algorithm software and engineering
The picture quality integrated treatment hoisting power of model machine hardware system, and more fully and effectively instruct the design of Processing Algorithm, test
Card and evaluation work.
（3）Image procossing lifting processor is replaced by using the combined type processing platform based on GPU, FPGA and multiple DSP
Change traditional CPU or DSP processing platforms and carry out algorithm acceleration.Global optimization and engineering are carried out from system level to algorithm, can
Completed to carry out the degraded image of resolution ratio as matter enhancing processing in 10 seconds.Hardware platform is by high-speed PCI E buses and everywhere
Plate connection is managed, parallel view data in equipment, in real time flowing water, processing can be achieved.
（4）The input of system, output interface use ten thousand mbit ethernet interfaces, and traffic rate reaches 10Gbps, and big chi can be achieved
The high-speed digital transmission of very little high-definition picture, the floor treatment for the high-resolution optical remote sensing satellite image that faces the future lifts examination
Test with engineering application value；System can be realized that image restoration and reconstruction, compression of images, image segmentation etc. are a variety of by user's request
Analysis, checking and the assessment of image processing algorithm, scalability are strong；Each building block in system uses optimal high property
Energy product, ensure that the Efficient Operation of system.
Brief description of the drawings
Fig. 1 is the Optical remote satellite image floor treatment lifting test system structure diagram of the present invention；
The Optical remote satellite image floor treatment lifting test system image procossing lifting processor composition that Fig. 2 is the present invention shows
It is intended to；
Fig. 3 is the Optical remote satellite image floor treatment lifting test system mixed filtering Processing Algorithm flow chart of the present invention.
Embodiment
Technical scheme is further described below in conjunction with the accompanying drawings, but is not limited to this, it is every to this hair
Bright technical scheme is modified or equivalent, without departing from the scope of the present invention program, all should be covered in the present invention
Protection domain among.
Embodiment one：As shown in figure 1, present embodiment record is at a kind of Optical remote satellite image ground
Lifting test system is managed, the pilot system includes data and sends work station, digital image procossing lifting work station, at image
Reason lifting processor, processing improving performance evaluation work station and Ethernet switch；The Ethernet switch is used to provide number
Processor, processing improving performance evaluation work are lifted according to work station, digital image procossing lifting work station, image procossing is sent
Ethernet connection between standing.
Need to carry out manual configuration to the IP address of all devices, to ensure that data transmission results in the IP of destination node
Address, sending node can find the path for reaching destination node by IP address.
Embodiment two：Optical remote satellite image floor treatment lifting test system described in embodiment one
System, the data, which send work station, to be used to receive original image and assistance data, and original image and assistance data are sent to totally
Word image procossing lifts work station and image procossing lifting processor is handled.
Data send work station in, user selection assistance data, including image sources, model type, product level, point
Resolution and scene type, configure corresponding management and control parameter, system performance parameter (MTF characteristics, noise characteristic, heterogeneity, quantization
Digit, compression algorithm, compression ratio etc.) and image-forming condition parameter (spot for photography, imaging time, shooting angle etc.), will be original
Image and auxiliary data transmission are to digital image procossing lifting work station, image procossing lifting processor.
Data input uses ten thousand mbit ethernet interfaces, and communication protocol is IEEE 802.3ae, and traffic rate is 10Gbps, is connect
Plug-in unit is SFP optical module (SFP).
Embodiment three：Optical remote satellite image floor treatment lifting test system described in embodiment one
System, the digital image procossing lifting work station is used to receive management and control data, original image and assistance data, carries out characteristic ginseng
Number is extracted, operation total variation image processing algorithm and mixed filtering image processing algorithm, and result is exported carried to processing
Rise Performance Evaluation work station.
If the digital image procossing lifting work station of management and control parameter selection operation, the work station is received from data transmission work
Original image and assistance data that station is sended over, judge whether there is transmission function and noise data in assistance data, if not having,
Carry out feature extraction.
Embodiment four：Optical remote satellite image floor treatment lifting test system described in embodiment one
System, described image processing lifting processor is used to receive management and control data, original image and assistance data, carries out characterisitic parameter and carries
Take, run what is accelerated based on graphic process unit GPU, on-site programmable gate array FPGA and multiple Digital Signal Processing DSP platforms
Total variation and mixed filtering combined type processing boosting algorithm, and result is exported give processing improving performance evaluation work station.
If management and control parameter selection operation image processing lifting processor, processor is received to be transmitted across from data transmission work station
The original image and assistance data come, judges whether include transmission function and noise data in assistance data, if not having, extracts and pass
Delivery function and noise data, extracting mode are consistent with digital image procossing lifting work station.It is based on according to these data runs
GPU, FPGA and multiple DSP total variation processing and the combined type integrated treatment boosting algorithm of mixed filtering.
Embodiment five：Optical remote satellite image floor treatment lifting test system described in embodiment one
System, the processing improving performance evaluation work station is used to receiving image, operation image matter after management and control instruction, original image and processing
Measure evaluation algorithms（Including image transmission function（MTF）Improving performance is evaluated, signal noise ratio (snr) of image（SNR）Improving performance is evaluated, image
The evaluation of quality metric (IQM) improving performance, image definition enhancing performance evaluation, image texture details retention property are evaluated, believed
Cease fidelity performance evaluation）, evaluated, provide final evaluation result.
Feature extraction includes image characteristics extraction and system performance is extracted.Image characteristics extraction is included based on recognition status
Transmission function is extracted, the transmission function based on natural image is extracted and noise characteristic is extracted；System performance, which is extracted, includes measurement
Error processing, measurement transfer function characteristics conversion, vibration characteristics to transmission function converted wave difference arrive transmission function conversion,
Noise characteristic is changed, and obtains transmission function and noise data.
According to these data run total variation Processing Algorithms and mixed filtering Processing Algorithm, and send result to treatability
Can evaluation work station.
Fig. 2 is that image procossing lifts processor composition schematic diagram, and the processor is blocked by two pieces of embedded image processing, two pieces
High-performance GPU processing card and an Intel Xeon high performance machine postures work station composition.
Wherein, the standard configuration of embedded image processing card is two pieces, supports additional extension increase, and every piece of processing card includes two
Piece DSP Processor and a piece of FPGA associations processing unit, the logic judgment complexity being collectively responsible in algorithm, processing procedure cache fragment
Change and the high part of requirement of real-time；High-performance GPU processing card standard configurations are two pieces, support additional extension increase, processing card
Using NVIDIA Tesla K80 GPU, it is responsible in algorithm in data set, large-scale Floating-point Computation；High-performance workstation uses
Intel Xeon processors, are managed, and pass through high-speed PCI E to input image data stream and each board handling process, parameter
Bus is connected with each process plate, is realized in equipment parallel view data, flowing water, is quasi real time handled.
Image output uses ten thousand mbit ethernet interfaces after processing, and communication protocol is IEEE 802.3ae, and traffic rate is
10Gbps, connector is SFP.
It is preferred that, by taking mixed filtering Processing Algorithm as an example, the flow of the algorithm is as follows：
1）Calculated using the Deconvolution Method based on Laplace regularization；
2）It is rightWavelet transformation is carried out, noisy wavelet coefficient is obtained；
3）Estimate wavelet in-band noise variance；
4）Thresholding, obtain wavelet coefficient after denoising；
5）It is rightInverse wavelet transform is carried out, final result is obtained。
The flow chart of mixed filtering Processing Algorithm is as shown in Figure 3.
Important step in flow is analyzed：
1. frequency filter is constructed：
The step is operated using pointwise.Due to being that pointwise is operated, algorithm can be assigned on parallel same processor core.Therefore
Dynamic range to data storage proposes higher requirement.In calculating process, major calculations type multiplies for continuous floating number
Method, addition and division, are adapted on FPGA, Floating-point DSP or GPU the operation of parallel and flowing water.
2. frequency domain filtering：
The step converts a signal into frequency domain using two-dimensional Fourier transform and carries out dot product, is turned result by inverse Fourier transform
Gain time domain.Two-dimensional Fourier transform first does one-dimensional Fourier transform line by line, then does one-dimensional Fourier transform by column.Therefore in Fu
In module can be divided inside leaf transformation, algorithm is first pressed into line splitting, every division calculates some row one-dimensional Fourier transforms,
Divide after result is collected according still further to row, every division calculates some one-dimensional Fourier transforms.Although this brings the number of complexity
According to transmitting-receiving with it is synchronous, but successfully algoritic module is divided so that module can divide and can run parallel.Based in two-dimentional Fu
The computing of leaf transformation and dot product, can be realized by way of multithreading, and parallel on a processor or increase is compared with complex data
Algoritic module is averagely divided and runs on other identical processors by control.
3. fast wavelet transform：
In real-time imaging system, wavelet transformation has larger amount of calculation, simple to meet real-time by software
Demand, therefore occur in that many new methods for realizing Wavelet Transformation Algorithm on a hardware platform.From hardware solution,
Common implementation has following several：
Based on general DSP hardware platform；
Hardware platform based on FPGA；
ASIP (ASIP) hardware platform based on the above two；
Based on unified calculation equipment framework CUDA platforms etc..
Here use and parallel accelerate is realized in FPGA platform.
Image procossing lifting processor is received from outside processing image to be restored and image processing algorithm parameter, according to
Above-mentioned steps operation mixed filtering Processing Algorithm is handled image, and result is sent into process performance evaluation work
Stand.
Process performance evaluation work station receives digital image procossing lifting work station or image procossing lifting processor hair
The image brought, the image after before processing is shown in the lump, and provides real-time the performance test results；Carry out property after before processing
The calculating of energy evaluation index, including image MTF improving performances evaluation algorithms, image SNR improving performances evaluation algorithms, image IQM
Improving performance evaluation algorithms, image definition enhancing Performance Evaluation Algorithm, image texture details retention property evaluation algorithms, information
Fidelity Performance Evaluation Algorithm etc., provides final Evaluation results.Finally by original image, image after processing, evaluation result
Send to outernal display unit and shown.
If evaluation is unsatisfactory for user's request, user can carry out image procossing again, by many by adjustment algorithm parameter
Secondary processing and evaluation iteration, optimize system process performance.
Embodiment six：Optical remote satellite image floor treatment lifting test system described in embodiment three
System, if the assistance data does not include transmission function and noise data, carries out image characteristics extraction and obtains transmission function and make an uproar
Sound data, operation total variation Processing Algorithm and mixed filtering Processing Algorithm, and result is exported commented to processing improving performance
Estimate work station.
Embodiment seven：Optical remote satellite image floor treatment lifting test system described in embodiment four
System, if the assistance data does not include transmission function and noise data, carries out image characteristics extraction and obtains transmission function and make an uproar
Sound data, run total variation processing and the combined type integrated treatment of mixed filtering lifting based on GPU, FPGA and multiple DSP and calculate
Method, and result is exported give processing improving performance evaluation work station.
Based in GPU, FPGA and multiple DSP combined types processing hardware platform, by the streamlined to handling process and simultaneously
Rowization optimizes, and allows GPU to be responsible in algorithm in data set, large-scale floating-point operation, and many DSP Processors and FPGA are responsible in algorithm
Logic judgment is complicated, processing procedure caching fragmentation and the high part of requirement of real-time.
Claims (7)
1. a kind of Optical remote satellite image floor treatment lifting test system, it is characterised in that：The pilot system includes number
Processor, processing improving performance evaluation work are lifted according to work station, digital image procossing lifting work station, image procossing is sent
Stand and Ethernet switch；The Ethernet switch is used to provide data transmission work station, digital image procossing lifting work
Stand, image procossing lifting processor, the Ethernet connection between processing improving performance evaluation work station.
2. Optical remote satellite image floor treatment lifting test system according to claim 1, it is characterised in that：It is described
Data, which send work station, to be used to receive original image and assistance data, and original image and assistance data are sent at digital image
Reason lifting work station and image procossing lifting processor are handled.
3. Optical remote satellite image floor treatment lifting test system according to claim 1, it is characterised in that：It is described
Digital image procossing lifting work station is used to receive management and control data, original image and assistance data, carries out characterisitic parameter extraction,
Total variation image processing algorithm and mixed filtering image processing algorithm are run, and result is exported commented to processing improving performance
Estimate work station.
4. Optical remote satellite image floor treatment lifting test system according to claim 1, it is characterised in that：It is described
Image procossing lifting processor is used to receive management and control data, original image and assistance data, carries out characterisitic parameter extraction, runs base
The total variation that accelerates in graphic process unit GPU, on-site programmable gate array FPGA and multiple Digital Signal Processing DSP platforms and mixed
Filtering combined type processing boosting algorithm is closed, and result is exported gives processing improving performance evaluation work station.
5. Optical remote satellite image floor treatment lifting test system according to claim 1, it is characterised in that：It is described
Processing improving performance evaluation work station is used to receiving image, operation image quality evaluation after management and control instruction, original image and processing
Algorithm, is evaluated, and provides final evaluation result.
6. Optical remote satellite image floor treatment lifting test system according to claim 3, it is characterised in that：If institute
Stating assistance data does not include transmission function and noise data, then carries out image characteristics extraction and obtain transmission function and noise data,
Total variation Processing Algorithm and mixed filtering Processing Algorithm are run, and result is exported gives processing improving performance evaluation work
Stand.
7. Optical remote satellite image floor treatment lifting test system according to claim 4, it is characterised in that：If institute
Stating assistance data does not include transmission function and noise data, then carries out image characteristics extraction and obtain transmission function and noise data,
The total variation processing based on GPU, FPGA and multiple DSP and the combined type integrated treatment boosting algorithm of mixed filtering are run, and will
Result, which is exported, gives processing improving performance evaluation work station.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710168902.XA CN106952255A (en) | 2017-03-21 | 2017-03-21 | Optical remote satellite image floor treatment lifting test system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710168902.XA CN106952255A (en) | 2017-03-21 | 2017-03-21 | Optical remote satellite image floor treatment lifting test system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106952255A true CN106952255A (en) | 2017-07-14 |
Family
ID=59473457
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710168902.XA Pending CN106952255A (en) | 2017-03-21 | 2017-03-21 | Optical remote satellite image floor treatment lifting test system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106952255A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107967198A (en) * | 2017-12-12 | 2018-04-27 | 华中科技大学 | A kind of star epigraph handles Simulation Evaluation device |
| CN111610001A (en) * | 2020-05-25 | 2020-09-01 | 中国科学院长春光学精密机械与物理研究所 | Wide remote sensing image MTF ground simulation testing device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101783008A (en) * | 2009-10-13 | 2010-07-21 | 上海海事大学 | Real-time processing platform for ultra high resolution remote sensing images based on functions of FPGA and DSP |
| CN101982976A (en) * | 2010-09-29 | 2011-03-02 | 中国科学院国家天文台 | System for displaying real-time data of linear-array stereo camera |
| CN103391428A (en) * | 2013-07-19 | 2013-11-13 | 中国民航大学 | Aviation meteorological image monitoring system based on multiple data links |
| CN104301627A (en) * | 2014-09-12 | 2015-01-21 | 上海卫星工程研究所 | Data pre-processing system and method of small wide-width ocean optics remote sensing satellite |
| CN205304979U (en) * | 2015-11-13 | 2016-06-08 | 河海大学 | Emergency satellite communication image transmission system |
-
2017
- 2017-03-21 CN CN201710168902.XA patent/CN106952255A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101783008A (en) * | 2009-10-13 | 2010-07-21 | 上海海事大学 | Real-time processing platform for ultra high resolution remote sensing images based on functions of FPGA and DSP |
| CN101982976A (en) * | 2010-09-29 | 2011-03-02 | 中国科学院国家天文台 | System for displaying real-time data of linear-array stereo camera |
| CN103391428A (en) * | 2013-07-19 | 2013-11-13 | 中国民航大学 | Aviation meteorological image monitoring system based on multiple data links |
| CN104301627A (en) * | 2014-09-12 | 2015-01-21 | 上海卫星工程研究所 | Data pre-processing system and method of small wide-width ocean optics remote sensing satellite |
| CN205304979U (en) * | 2015-11-13 | 2016-06-08 | 河海大学 | Emergency satellite communication image transmission system |
Non-Patent Citations (3)
| Title |
|---|
| 中国科学院上海技术物理研究所著: "《红外光电技术》", 30 March 2004, 中国科学技术出版社 * |
| 崔光茫: "光学遥感图像质量提升及评价技术研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
| 王明远等: "亚洲一流的遥感卫星地面应用系统", 《空间电子技术》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107967198A (en) * | 2017-12-12 | 2018-04-27 | 华中科技大学 | A kind of star epigraph handles Simulation Evaluation device |
| CN111610001A (en) * | 2020-05-25 | 2020-09-01 | 中国科学院长春光学精密机械与物理研究所 | Wide remote sensing image MTF ground simulation testing device |
| CN111610001B (en) * | 2020-05-25 | 2021-09-21 | 中国科学院长春光学精密机械与物理研究所 | Wide remote sensing image MTF ground simulation testing device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113222822B (en) | Hyperspectral image super-resolution reconstruction method based on multi-scale transformation | |
| Shao et al. | Branchy-GNN: A device-edge co-inference framework for efficient point cloud processing | |
| CN105976383A (en) | Power transmission equipment fault diagnosis method based on limit learning machine image recognition | |
| CN113192646B (en) | Target detection model construction method and device for monitoring distance between different targets | |
| CN110288526B (en) | Optimization method for improving imaging quality of single-pixel camera by image reconstruction algorithm based on deep learning | |
| CN108460749B (en) | Rapid fusion method of hyperspectral and multispectral images | |
| CN102148987A (en) | Compressed sensing image reconstructing method based on prior model and 10 norms | |
| CN106952255A (en) | Optical remote satellite image floor treatment lifting test system | |
| CN117876397B (en) | Bridge member three-dimensional point cloud segmentation method based on multi-view data fusion | |
| CN114972323B (en) | Defect picture generation method and system for model training | |
| CN104715470B (en) | A kind of klt Corner Detections device and method | |
| CN116128820A (en) | Pin state identification method based on improved YOLO model | |
| CN116416491A (en) | Remote sensing pseudo sample generation method based on lightweight diffusion model | |
| CN104159048A (en) | Compressive sensing uniform weighting relevance imaging method for non-uniform light field | |
| CN113096015B (en) | Image super-resolution reconstruction method based on progressive perception and ultra-lightweight network | |
| CN110728352A (en) | Large-scale image classification method based on deep convolutional neural network | |
| CN107680126B (en) | Random sampling consistency image matching denoising processing system and method | |
| CN114155252A (en) | Human body target segmentation method and system for mobile terminal application | |
| CN109658467A (en) | A kind of endoscopic images sensing reconstructing method based on multiword allusion quotation modified compressed sensing framework | |
| CN115035408A (en) | Unmanned aerial vehicle image tree species classification method based on transfer learning and attention mechanism | |
| CN111260570B (en) | Binarization background noise simulation method for posts based on cyclic consistency confrontation network | |
| CN114494138A (en) | Photovoltaic panel defect detection method and system | |
| Nourian et al. | Image de-noising with virtual hexagonal image structure | |
| Yang et al. | Improving Semantic Segmentation Performance by Jointly Using High Resolution Remote Sensing Image and Ndsm | |
| CN113658076B (en) | Image restoration method, device, equipment and medium based on feature entanglement modulation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170714 |
|
| RJ01 | Rejection of invention patent application after publication |