CN106886786A - A kind of effective image processing system - Google Patents

A kind of effective image processing system Download PDF

Info

Publication number
CN106886786A
CN106886786A CN201710104807.3A CN201710104807A CN106886786A CN 106886786 A CN106886786 A CN 106886786A CN 201710104807 A CN201710104807 A CN 201710104807A CN 106886786 A CN106886786 A CN 106886786A
Authority
CN
China
Prior art keywords
image
module
image processing
histogram
pixel
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
Application number
CN201710104807.3A
Other languages
Chinese (zh)
Inventor
不公告发明人
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Xun Ye Computer Technology Co Ltd
Original Assignee
Shanghai Xun Ye Computer Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanghai Xun Ye Computer Technology Co Ltd filed Critical Shanghai Xun Ye Computer Technology Co Ltd
Priority to CN201710104807.3A priority Critical patent/CN106886786A/en
Publication of CN106886786A publication Critical patent/CN106886786A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/211—Selection of the most significant subset of features
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/56—Extraction of image or video features relating to colour

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Multimedia (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Image Processing (AREA)
  • Image Analysis (AREA)

Abstract

A kind of effective image processing system, including image capture module, image transmission module, image processing module and image storage module；Described image acquisition module is used to collect original image, and image processing module is used to carry out the original image transmitted through image transmission module subsequent treatment, the image that image storage module is used for after storage treatment.Beneficial effects of the present invention are：Setting image capture module, image transmission module, image processing module and image storage module, simple structure, flexibly and easily；And by the mutual cooperation between each module, the original image that will timely and effectively can be collected is transferred to image processing module exactly, image processing module carries out the Multilevel method of shadow removal, Objective extraction and histogram inspection to image, it is ensured that the accuracy of target detection；Image storage module is stored to the image after treatment, is facilitated later inquiry and is used.

Description

A kind of effective image processing system
Technical field
The invention is related to technical field of image processing, and in particular to a kind of effective image processing system.
Background technology
People from the most information that the external world obtains are obtained from image, image are carried out using computer various The treatment of form, the development of image processing techniques is promoted so as to the mode for obtaining the information that image is included.
In the image processing arts, how accurately to detect that target is a big key issue of image processing techniques, but During target detection, often due to the influence of scene changes and shade etc., the accuracy to target detection brings pole Big difficulty, thus propose it is a kind of can effective detection had great significance to the image processing system of target.
The content of the invention
Regarding to the issue above, the present invention is intended to provide a kind of effective image processing system.
The purpose of the invention is achieved through the following technical solutions：
A kind of effective image processing system, including image capture module, image transmission module, image processing module and figure As memory module；Described image acquisition module is used to collect original image, and image processing module is used for through image transmission module The original image of transmission carries out subsequent treatment, the image that image storage module is used for after storage treatment.
Preferably, described image acquisition module uses camera collection image.
Preferably, image pre-processing module is additionally provided with, described image pretreatment module is connected with image transmission module, is used Pre-processed in the original image.
Preferably, described image processing module also includes shadow removal unit, Objective extraction unit and histogram checklist Unit.
Preferably, the shadow removal unit is used for from the interval histogram of the extracting target from images for receiving and removes The shade of target interval, specifically includes：
A. R, G, B color value in acquisition image corresponding to each pixel i, set up color spaceSpecially：
In formula, N is the sum of pixel in image, ri、gi、biR, G, B color component respectively corresponding to pixel i；
B. defining matrix U is used for 3-dimensional vectorDimension-reduction treatment is carried out, is specifically included：
In formula,
C. the bivector x that will be obtainediProjection process is carried out, the gray value I after being projectedi；
D. the assertive evidence gray value I that will be obtainediProcessed according to the following formula, specially：
E. gray value I ' is set upiHistogram, take a fixed packet count z is expired with obtaining fixed group away from, packet count z Sufficient minimum packets number；
F. histogram probability g is calculatedj, in the hope of intrinsic figure angle, θ entropy, its computing formula is：
In formula, djIt is the pixel number in set of histograms j, gjIt is pixel probability in set of histograms j, z is grouped for histogram Number, ρθIt is histogram information entropy；
G. using being projected from 0 to 180 °, the intrinsic figure that each Angles Projections is obtained is obtained, is calculated according to the method described above Intrinsic figure each angle, θ entropy, so as to try to achieve minimum entropy ρθ′, specially：
ρθ′=min ρθθ∈(0,180°)
H. minimum entropy ρθ′Corresponding angle is exactly intrinsic angle θ ', the unrelated figure of illumination that its corresponding grey scale image is target So。
Beneficial effects of the present invention are：Image capture module, image transmission module, image processing module and image is set to deposit Storage module, simple structure, flexibly and easily；By the mutual cooperation between each module, the original that will timely and effectively can be collected Beginning image is transferred to image processing module exactly, and image processing module carries out shadow removal, Objective extraction and straight to image The Multilevel method of side's figure inspection, it is ensured that the accuracy of target detection；Image storage module is stored to the image after treatment, Facilitate later inquiry and use.
Brief description of the drawings
Innovation and creation are described further using accompanying drawing, but embodiment in accompanying drawing does not constitute and the invention is appointed What is limited, for one of ordinary skill in the art, on the premise of not paying creative work, can also be according to the following drawings Obtain other accompanying drawings.
Fig. 1 is schematic structural view of the invention；
Fig. 2 is the structural representation of image pre-processing module；
Fig. 3 is the structural representation of image processing module.
Reference：
Image capture module 1, image pre-processing module 2, image transmission module 3, image processing module 4, image are deposited Storage module 5, image culling unit 21, image cropping unit 22, image encryption unit 23, shadow removal unit 41, target Extraction unit 42, histogram verification unit 43.
Specific embodiment
The invention will be further described with the following Examples.
Referring to Fig. 1, a kind of effective image processing system of the present embodiment, including image capture module 1, image transmitting mould Block 3, image processing module 4 and image storage module 5；Described image acquisition module 1 is used to collect original image, image procossing mould Block 4 is used to carry out the original image transmitted through image transmission module 3 subsequent treatment, and image storage module 5 is processed for storage Image afterwards.
Preferably, described image acquisition module 1 uses camera collection image.
The above embodiment of the present invention sets image capture module 1, image transmission module 3, image processing module 4 and image and deposits Storage module 5, simple structure, flexibly and easily；And by the mutual cooperation between each module, will timely and effectively can collect Original image be transferred to image processing module 4 exactly, image processing module 4 carries out shadow removal, Objective extraction to image And the Multilevel method of histogram inspection, it is ensured that the accuracy of target detection；Image after 5 pairs for the treatment of of image storage module enters Row storage, facilitates later inquiry and uses.
Preferably, image pre-processing module 2 is additionally provided with, described image pretreatment module 2 connects with image transmission module 3 Connect, for being pre-processed to the original image, described image pretreatment module 2 is cut out including image culling unit 21, image Unit 22, image encryption unit 23 are cut, wherein：
Described image culling unit 21 is used to enter the redundant image and the too poor image of quality of the collection of data acquisition module 1 Row is rejected；
The image that described image cuts unit 22 and is used for after being rejected to image culling unit 21 cuts；
The image that described image ciphering unit is used for after being cut to image cropping unit 22 is encrypted；
This preferred embodiment constructs image preprocessing system 2, realizes to eliminating redundant image, and meter is greatly reduced Calculation amount, improves image processing efficiency；Effective cutting to residual image after rejecting, solves the figure that sensor lenses are brought Image distortion problem；Image after cutting is encrypted, it is ensured that reliability and security of the image in transmitting procedure, overcome Conventional encryption algorithm traffic load is big to wait not enough.
Preferably, described image culling unit 21 is rejected for the image too small to otherness, and the image difference opposite sex is adopted Judged with following manner：
A, for two images A (x, y) and B (p, q), for the pixel of same position on image, define otherness public Formula,
In formula, S (A, B) represents the otherness of two images, and R (x, y), G (x, y), B (x, y) are represented in A images respectively Red, green and blue color component value, R (p, q), G (p, q), B (p, q) represent red, the green and indigo plant in B images respectively Color color component value；
If pixel of the similitude less than 0.12 is more than 10% in b, image, wherein piece image is randomly selected as weight Complex pattern is rejected.
This preferred embodiment eliminates multiimage, and amount of calculation is greatly reduced, and improves image processing efficiency.
Preferably, described image cuts the image after unit 22 pairs is rejected and cuts, and specifically includes：
If image size is M × N, the edge to image cuts, and retains picture centre region, the shared figure in central area As area ratio is determined using equation below：
In formula, r represents that the image after cutting accounts for original image area ratio, FminIt is sensor shortest focal length, F is adopted for sensor Real focal length used during collection image；
This preferred embodiment solves the problems, such as the pattern distortion that sensor lenses are brought.
Preferably, described image ciphering unit 23 is used to be encrypted the image after cutting, including：
A, set cut after the size of coloured image be P × Q, extraction rgb color component；
B, coloured image is divided into multiple size identical grid, each grid is entered using long-lost cosine code then Line translation, obtains each color component coefficient matrix；
C, to each color component coefficient using being transmitted after des encryption algorithm for encryption.
This preferred embodiment ensure that reliability and security of the image in transmitting procedure, overcome conventional encryption algorithm Traffic load is big to wait not enough.
Preferably, described image processing module 4 also includes the inspection of shadow removal unit 41, Objective extraction unit 42 and histogram Verification certificate unit 43, wherein：
The shadow removal unit 41 is used for from the interval histogram of the extracting target from images for receiving and removes target Interval shade；
The Objective extraction unit 42 is used for the histogram of the target interval being partitioned into from the image for receiving in image The interval histogram with shade；
The histogram verification unit 43 is used to check the histogram of shadow removal unit 41 and Objective extraction unit 42, from And obtain target interval；
This preferred embodiment constructs image processing system 4, realizes the effective treatment to the image for receiving, and improves The verification and measurement ratio of target.
Preferably, the shadow removal module 41 is used to be obtained from the image for receiving the unrelated figure of illumination of target, tool Body includes：
A. R, G, B color value in acquisition image corresponding to each pixel i, set up color spaceSpecially：
In formula, N is the sum of pixel in image, ri、gi、biR, G, B color value respectively corresponding to pixel i.
B. defining matrix U is used for 3-dimensional vectorDimension-reduction treatment is carried out, is specifically included：
In formula,
C. the bivector x that will be obtainediProjection process is carried out, the gray value I after being projectedi, specially：
D. the assertive evidence gray value I that will be obtainediProcessed according to the following formula, specially：
E. gray value I ' is set upiHistogram, take a fixed packet count z is expired with obtaining fixed group away from, packet count z Sufficient minimum packets number；
F. histogram probability g is calculatedj, in the hope of intrinsic figure angle, θ entropy, its computing formula is：
In formula, djIt is the pixel number in set of histograms j, gjIt is pixel probability in set of histograms j, z is grouped for histogram Number, ρθIt is histogram information entropy；
G. using being projected from 0 to 180 °, the intrinsic figure that each Angles Projections is obtained is obtained, is calculated according to the method described above Intrinsic figure each angle, θ entropy, so as to try to achieve minimum entropy ρθ′, specially：
ρθ′=min ρθθ∈(0,180°)
H. minimum entropy ρθ′Corresponding angle is exactly intrinsic angle θ ', the unrelated figure of illumination that its corresponding grey scale image is target So。
The shadow removal module 41 that this preferred embodiment is set, compared with prior art, using fixation when histogram is set up Group away from overcoming set of histograms and calculate the instability problem that brings to comentropy away from number change；According to the unrelated figure of illumination Carry out target detection, it is ensured that the verification and measurement ratio of target.
Preferably, the Objective extraction unit 42, in the base containing hypographous target S obtained using background subtraction The interval histogram S of the histogram and shade of target interval is partitioned on plinthi(i=1,2), specifically includes：
A. rgb color space model is set up, the R corresponding to each pixel is obtainedi、Gi、BiComponent；
B. the corresponding gray value H of each pixel is calculated according to following equationi；
Hi=α Ri+βGi+γBi(i=1,2 ... M)
Wherein, α, β, γ are respectively Ri、Gi、BiComponent meets α, β, γ >=0, alpha+beta+γ=1 to the weights of gray value, M is pixel sum.
C. the intrinsic gray value H that will be obtainediProcessed according to the following formula；
D. according to gray value HiThe grey level histogram of image is set up, its packet count is determined for k, its cut-point is k；
E. each cut-point is set to initial threshold, and the histogram of each threshold value the right and left is counted using following equation Calculate；
Wherein, n=1,2 ... k, k are histogram packet count, miIt is the pixel count in group i, n is initial threshold；
F. the above-mentioned h being calculatedn' corresponding initial threshold n is final threshold value τ；
G., the histogram divion of target T can be obtained histogram and the shadow region of target interval according to the threshold tau for obtaining Between histogram, be utilized respectively statistics with histogram target interval and shade interval in object area：
Wherein, miIt is the pixel number in set of histograms i, piFor image gray levels are the probability of the pixel of i, S1、S2Respectively It is the area of object in histogram.
The object extraction module 42 that above-described embodiment is set, compared with prior art, transition is not between shade and target When substantially, the interval histogram of the histogram and shade of target interval is preferably partitioned into, improves the verification and measurement ratio of target interval；
Preferably, the histogram inspection module 43 is used to check the straight of shadow removal unit 41 and Objective extraction unit 42 Fang Tu, so as to obtain target interval, specifically includes：
A. the unrelated figure S of illumination is calculatedoHistogram in object area, computing formula is：
Wherein, N is histogram pixel point sum, giFor image gray levels are the probability of the pixel of i, S0It is thing in histogram The area of body.
B. target interval S is definediInspection formula it is as follows：
Si=| Si-So|-f (i=1,2)
Si<When 0, SiAs target interval, Si>When 0, SiAs shade is interval, and f is detection threshold value, and 0.05 is set to here.
Inventor has done a series of tests for the present embodiment, and test result is as shown in the table：
Detection target Target recall rate
Fixed object 100%
Mobiles 100%
Fixing human 100%
Moving human body 99%
Mobiles 99%
As can be seen that by the image of the system treatment, target recall rate is maintained at more than 98% from test data.
The histogram verification unit 43 that this preferred embodiment is set, compared with prior art, using shadow removal unit 41 The histogram of the target interval of acquisition as inspection foundation, the histogram of the target interval obtained with Objective extraction unit 42 and The interval histogram of shade is tested, it is ensured that the accuracy of target detection.
Finally it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than the present invention is protected The limitation of scope is protected, although being explained to the present invention with reference to preferred embodiment, one of ordinary skill in the art should Work as understanding, technical scheme can be modified or equivalent, without deviating from the reality of technical solution of the present invention Matter and scope.

Claims (5)

1. a kind of effective image processing system, it is characterized in that, including image capture module, image transmission module, image procossing Module and image storage module；Described image acquisition module is used to collect original image, and image processing module is used for through image The original image of transport module transmission carries out subsequent treatment, the data that image storage module is used for after storage treatment.
2. a kind of effective image processing system according to claim 1, it is characterized in that, described image acquisition module is used Camera collection image.
3. a kind of effective image processing system according to claim 2, it is characterized in that, it is additionally provided with image preprocessing mould Block, described image pretreatment module is connected with image transmission module, for being pre-processed to the original image.
4. a kind of effective image processing system according to claim 1, it is characterized in that, described image processing module is also wrapped Include shadow removal unit, Objective extraction unit and histogram verification unit.
5. a kind of effective image processing system according to claim 4, it is characterized in that, the shadow removal unit is used for From the interval histogram of the extracting target from images that receives and the shade of target interval is removed, specifically included：
A. R, G, B color value in acquisition image corresponding to each pixel i, set up color spaceSpecially：
In formula, N is the sum of pixel in image, ri、gi、biR, G, B color value respectively corresponding to pixel i；
B. defining matrix U is used for 3-dimensional vectorDimension-reduction treatment is carried out, is specifically included：
In formula,
C. the bivector x that will be obtainediProjection process is carried out, the gray value I after being projectedi, specially：
I i = ( c o s &theta; , s i n &theta; ) m i n i = m i c o s &theta; + n i s i n &theta; ( i = 1 , 2 ... ... N )
D. the assertive evidence gray value of acquisition is processed according to the following formula, specially：
I &prime; i = I i max N I i + min N I i
E. gray value I ' is set upiHistogram, take a fixed packet count z and met most away from, packet count z with obtaining fixed group Small packet count；
z &GreaterEqual; 2 N 3 = z m i n
F. histogram probability g is calculatedj, in the hope of intrinsic figure angle, θ entropy, its computing formula is：
g j = d j &Sigma; j = 1 z d j
&rho; &theta; = - &Sigma; j z g j log 2 g j
In formula, djIt is the pixel number in set of histograms j, gjIt is pixel probability in set of histograms j, z is histogram packet count, ρθIt is histogram information entropy；
G. using being projected to 180 degree from 0, the intrinsic figure that each Angles Projections is obtained is obtained, this is calculated according to the method described above Entropy of the figure in each angle, θ is levied, so as to try to achieve minimum entropy ρθ′, specially：
ρθ′=min ρθ
H. minimum entropy ρθ′Corresponding angle is exactly intrinsic angle θ ', the unrelated figure S of illumination that its corresponding grey scale image is targeto。
CN201710104807.3A 2017-02-24 2017-02-24 A kind of effective image processing system Pending CN106886786A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710104807.3A CN106886786A (en) 2017-02-24 2017-02-24 A kind of effective image processing system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710104807.3A CN106886786A (en) 2017-02-24 2017-02-24 A kind of effective image processing system

Publications (1)

Publication Number Publication Date
CN106886786A true CN106886786A (en) 2017-06-23

Family

ID=59180204

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710104807.3A Pending CN106886786A (en) 2017-02-24 2017-02-24 A kind of effective image processing system

Country Status (1)

Country Link
CN (1) CN106886786A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107729950A (en) * 2017-11-07 2018-02-23 钟永松 A kind of good district monitoring system of monitoring effect
CN117726923A (en) * 2024-02-05 2024-03-19 河北凡谷科技有限公司 Image communication system based on specific model

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060285769A1 (en) * 2005-06-20 2006-12-21 Samsung Electronics Co., Ltd. Method, apparatus, and medium for removing shading of image
CN101556739A (en) * 2009-05-14 2009-10-14 浙江大学 Vehicle detecting algorithm based on intrinsic image decomposition
US20100253684A1 (en) * 2009-04-02 2010-10-07 Qingxiong Yang Methods for obtaining a three-band intrinsic image from a single input image
GB2528994A (en) * 2014-08-08 2016-02-10 Imagination Tech Ltd Relightable texture for use in rendering an image
CN105453543A (en) * 2013-08-07 2016-03-30 高通股份有限公司 Dynamic color shading correction
US20160140722A1 (en) * 2014-11-18 2016-05-19 Adobe Systems Incorporated Fast Intrinsic Images

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20060285769A1 (en) * 2005-06-20 2006-12-21 Samsung Electronics Co., Ltd. Method, apparatus, and medium for removing shading of image
US20100253684A1 (en) * 2009-04-02 2010-10-07 Qingxiong Yang Methods for obtaining a three-band intrinsic image from a single input image
CN101556739A (en) * 2009-05-14 2009-10-14 浙江大学 Vehicle detecting algorithm based on intrinsic image decomposition
CN105453543A (en) * 2013-08-07 2016-03-30 高通股份有限公司 Dynamic color shading correction
GB2528994A (en) * 2014-08-08 2016-02-10 Imagination Tech Ltd Relightable texture for use in rendering an image
US20160140722A1 (en) * 2014-11-18 2016-05-19 Adobe Systems Incorporated Fast Intrinsic Images

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
郭玲: "智能视频监控中运动目标检测的算法研究", 《中国博士学位论文全文数据库 信息科技辑》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107729950A (en) * 2017-11-07 2018-02-23 钟永松 A kind of good district monitoring system of monitoring effect
CN117726923A (en) * 2024-02-05 2024-03-19 河北凡谷科技有限公司 Image communication system based on specific model
CN117726923B (en) * 2024-02-05 2024-05-14 河北凡谷科技有限公司 Image communication system based on specific model

Similar Documents

Publication Publication Date Title
CN105719188B (en) The anti-method cheated of settlement of insurance claim and server are realized based on plurality of pictures uniformity
CN110991266B (en) Binocular face living body detection method and device
CN104700404B (en) A kind of fruit positioning identifying method
CN105279372B (en) A kind of method and apparatus of determining depth of building
CN107481343A (en) A kind of check class attendance based on face recognition technology is registered system and its method of work
CN106709950A (en) Binocular-vision-based cross-obstacle lead positioning method of line patrol robot
CN106407927A (en) Salient visual method based on polarization imaging and applicable to underwater target detection
CN115272271A (en) Pipeline defect detecting and positioning ranging system based on binocular stereo vision
CN107665361A (en) A kind of passenger flow counting method based on recognition of face
CN111723656B (en) Smog detection method and device based on YOLO v3 and self-optimization
CN106886786A (en) A kind of effective image processing system
CN107527095A (en) A kind of vehicle detection image-pickup method and system
CN108471497A (en) A kind of ship target real-time detection method based on monopod video camera
CN114820758A (en) Plant growth height measuring method, device, electronic device and medium
CN104573647B (en) A kind of recognition methods of illegal identity certificate and device
CN110378351A (en) Seal discrimination method and device
CN114648669A (en) Motor train unit fault detection method and system based on domain-adaptive binocular parallax calculation
CN114091602A (en) SSR flow identification system and method based on machine learning
CN116320292A (en) Water conservancy monitoring control system based on big data
CN106331618B (en) A kind of method and device automatically confirming that video camera visible range
CN105469413B (en) It is a kind of based on normalization ring weighting without refer to smear restoration image synthesis method for evaluating quality
CN111178503A (en) Mobile terminal-oriented decentralized target detection model training method and system
CN113052110B (en) Three-dimensional interest point extraction method based on multi-view projection and deep learning
CN106600905A (en) Effective geological disaster monitoring system
CN106789485B (en) A kind of easily smart home monitoring 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
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20170623