CN110516686B - Mosquito recognition method of three-color RGB image - Google Patents

Mosquito recognition method of three-color RGB image Download PDF

Info

Publication number
CN110516686B
CN110516686B CN201910626708.0A CN201910626708A CN110516686B CN 110516686 B CN110516686 B CN 110516686B CN 201910626708 A CN201910626708 A CN 201910626708A CN 110516686 B CN110516686 B CN 110516686B
Authority
CN
China
Prior art keywords
mosquito
target
image
rgb
function
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.)
Active
Application number
CN201910626708.0A
Other languages
Chinese (zh)
Other versions
CN110516686A (en
Inventor
谢雪梅
苗宏达
吴家骥
杨众杰
谭铭洲
李甫
付博勋
景易星
金星
韩笑
杜曜辛
李旭超
杨文哲
徐显梁
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.)
Xidian University
Original Assignee
Xidian University
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 Xidian University filed Critical Xidian University
Priority to CN201910626708.0A priority Critical patent/CN110516686B/en
Publication of CN110516686A publication Critical patent/CN110516686A/en
Application granted granted Critical
Publication of CN110516686B publication Critical patent/CN110516686B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
    • 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
    • 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)
  • Multimedia (AREA)
  • Data Mining & Analysis (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a mosquito identification method of a three-color RGB image, which mainly solves the problems of more computing resources and high equipment cost in the prior art. The implementation scheme is as follows: acquiring an image area of a target mosquito, and traversing pixel points of a target mosquito image to obtain an RGB (red, green and blue) color histogram and a histogram function of the target mosquito image; designing RGB characteristic functions of different types of mosquito target images according to three-color RGB image histograms of the different types of mosquito target images; simplifying and designing the function expression according to the characteristics of the specific mosquito species and obtaining the parameters of the function; and classifying and identifying the target mosquitoes according to the RGB color histogram function of the target mosquitoes and the RGB characteristic function of the mosquitoes. The invention reduces the calculation resources, reduces the equipment cost, can realize the classification of two mosquitoes, namely culex pipiens pallens and aedes albopictus, and can be used for identifying biological species.

Description

Mosquito recognition method of three-color RGB image
Technical Field
The invention belongs to the technical field of digital image processing, and particularly relates to mosquito recognition in a picture, which can be used for recognizing biological species or objects.
Background
The traditional mosquito classification generally adopts an artificial method, namely, the collected mosquito samples are classified by naked eyes according to the morphological characteristics of the mosquitoes, and the defects of low classification efficiency, high risk and great influence by artificial characteristics such as fatigue exist. The method for processing the digital image can greatly improve the efficiency of mosquito identification and reduce the labor cost and the artificial risk.
With the development of the computer and intelligent fields, the target detection algorithm also makes continuous breakthrough. The traditional target detection algorithm is mainly used for detecting based on image features, for example, a Scale Invariant Feature Transform (SIFT) algorithm proposed by Lowe matches targets in an image by searching for feature points which are not easily subjected to illumination, noise, affine transformation and the like in the image, detects extreme points by using a Gaussian difference function, eliminates unstable extreme points, and finally obtains the features of a specified target by using a direction histogram to count the gradient of key points. The SIFT algorithm has the characteristics of good feature extraction and high robustness. But has the problems of high algorithm complexity and low detection speed.
In view of the disadvantages of SIFT, Yan Ke et al and Bay et al propose the PCA-SIFT method and SURF method, respectively. The PCA-SIFT algorithm introduces a principal component analysis method on the basis of SIFT, PCA is used for replacing a histogram to perform dimension reduction processing on the sub-vectors, and matching efficiency is improved. The SUFT introduces a Hessian matrix to obtain the key point positioning, and the calculation amount is greatly reduced. The Viola-Jones algorithm proposed by Paul Viola and Michael J Jones is the first universal face detection algorithm that can process in real time and has good effect. The Viola-Jones algorithm uses Haar features to describe the light and shade changes of a picture window and a local area, and uses a cascade classifier to reduce the operation amount, thereby realizing real-time target detection.
The existing target detection mainly utilizes deep learning to detect, and has two types of detection algorithms, one is a target detection algorithm based on classification, and the other is an algorithm for converting the classification into a regression problem. The first type mainly finds out a candidate region in the image through an OverFeat and R-CNN algorithm, and classifies the candidate region. The second type is a target detection algorithm mainly represented by YOLO and SSD, which predicts the region of the target in a regression manner and measures the probability of the class with confidence to complete the detection and identification of the target.
Because different mosquitoes have similar morphological characteristics, the same type of mosquito can exhibit different postures, which has certain difficulty in SIFT-based feature point detection. Although the algorithm based on deep learning has high detection accuracy, a large amount of computing resources are needed, and relatively strong computing equipment such as GPU equipment is needed, so that the cost is relatively high, and the requirement of widely using mosquito detection equipment in a relatively large region cannot be met.
Disclosure of Invention
The invention aims to provide a mosquito identification method of a three-color RGB image aiming at the defects of the prior art, so as to reduce computing resources, reduce equipment cost and meet the requirement of mosquito detection in a larger region.
The technical scheme of the invention is that the mosquito target contained in the picture to be detected is extracted by the characteristic that the picture to be detected and the background of the picture have difference; and extracting the color characteristics of the mosquito target area to obtain an RGB color histogram of the mosquito target area. And designing a characteristic function according to the statistical characteristics of the histogram and calculating confidence scores of different types of mosquitoes. And obtaining the recognition condition of the mosquitoes according to the confidence scores. The concrete implementation steps comprise:
(1) acquiring a single mosquito target image;
(2) traversing pixel points of the mosquito target image to obtain a three-color RGB image histogram of each mosquito target image and a three-color RGB image histogram function H (i, c), wherein c belongs to (R, G, B) and represents the category of three-color RGB, and i belongs to [1, 256] and represents the value of three-color RGB;
(3) designing RGB characteristic functions of different types of mosquito target images according to three-color RGB image histograms of the different types of mosquito target images:
Figure BDA0002127343410000021
wherein [ mosquito species]Is a set of mosquito species, m is a type of mosquito species set, ScoremFunction g representing confidence of the mosquito to be detected under m-type conditionsm(i, c) and function fm(i, c) respectively representing weighted weights of two different forms of histogram functions h (i, c) of the mosquito image to be detected under the m types of conditions;
(4) simplifying the design of the function expression according to the characteristics of the specific mosquito species and obtaining the parameters of the function:
(4a) weighting weights for two different forms of an histogram function h (i, c) under two mosquito conditions of culex pipiens pallens and aedes albopictus:
Figure BDA0002127343410000022
wherein Setmc＝[1，left]∪[right，256]A Set of RGB values is obtained, and left is SetmcRight of the subset
Boundary, right is SetmcThe left boundary of the subset;
(4b) setting the Score as the confidence coefficient of the mosquito to be detected under the condition of only aiming at two types of mosquitoes, namely culex pipiens pallens and aedes albopictus, and combining the function fm(i，c)、gm(i, c) substituting the RGB feature function in (3), thereby obtaining a simplified RGB feature function as follows:
Figure BDA0002127343410000031
(4c) setting a threshold value of confidence coefficient of the mosquitoes to be detected under the conditions of the culex pallens and the aedes albopictus as thresh, comparing the Score with the thresh, and distinguishing the types of the mosquitoes to be detected: if the Score is more than thresh, the target mosquito is Aedes albopictus, otherwise, the target mosquito is Culex pipiens pallens;
(4d) giving a group of RGB histogram functions of known culex pallens and aedes albopictus images, traversing all three parameters of left, right and thresh, sequentially substituting the RGB histogram function and the left of a given known mosquito image into the RGB characteristic function in each traversal, taking the right as a value of the traversal, calculating to obtain the value of the RGB characteristic function as Score, comparing the Score with the current value of the thresh to obtain a classification result of the round, obtaining the accuracy of the classification result under the current parameter through statistics after each traversal is finished, and obtaining the parameter with the highest accuracy of the mosquito classification result as an optimal parameter after all traversals are finished: left ', right ', thresh ';
(4e) substituting the optimal parameters left 'and right' obtained in the step (4d) into the RGB characteristic function obtained in the step (4b) to obtain a final characteristic function;
(5) calculating the confidence coefficient of the target mosquitoes and classifying the target mosquitoes:
substituting the histogram function H (i, c) of the target mosquito image obtained in the step (2) into the final characteristic function obtained in the step (4e), calculating the confidence Score ' of the target mosquito image under the condition of only aiming at two types of mosquitoes, namely culex pallens and aedes albopictus, comparing the Score ' value with the threshold value thresh ' obtained in the step (4d), and judging the type of the target mosquito: if Score 'is greater than thresh', the target mosquito is Aedes albopictus, otherwise, Culex pallens.
Compared with the prior art, the invention has the following advantages:
firstly, the invention extracts the RGB histogram characteristics of the image target and designs the characteristic function to carry out the identification and classification of the mosquitoes, thereby reducing the calculation complexity and realizing the real-time detection of the mosquitoes on the hardware equipment with low cost.
Secondly, the invention calculates the parameters of the RGB characteristic extraction function by using a method for extracting data characteristics in given mosquito image data, and calculates the confidence degrees of different types of mosquitoes according to the detected target mosquito image data of different types to obtain the category of the target mosquitoes.
Drawings
FIG. 1 is a flow chart of an implementation of the present invention.
Detailed description of the preferred embodiments
Examples of the present invention are described in further detail below with reference to the accompanying drawings.
Referring to fig. 1, the implementation steps for the example are as follows:
step 1, obtaining a single mosquito target image.
1.1) selecting a background picture, comparing the background picture with a target image to be detected by a background subtraction method, and acquiring a distribution area of mosquito targets in the image;
1.2) in a distribution area of the mosquito target, obtaining foreground image information by a background difference method, eliminating image noise by corrosion operation, and eliminating the influence of mosquito textures by expansion operation to further obtain a mosquito distribution area image;
1.3) respectively calculating the communication area of each mosquito target on the mosquito distribution area image to obtain a single mosquito target area image.
And 2, acquiring RGB statistical characteristics.
Traversing pixel points of the mosquito target image, calculating to obtain a three-color RGB image distribution histogram H (c, i) of the mosquito target image,
wherein c ∈ (R, G, B) represents the category of three-color RGB, i ∈ [1, 256] represents the value of three-color RGB, R represents red, G represents green, and B represents blue.
And 3, designing RGB characteristic functions of target images of different types of mosquitoes.
3.1) designing the original characteristic functions of RGB:
3.1.1) setting a set [ mosquitos ] as a mosquito kind set, wherein m is one kind of mosquito kind set, and obtaining a histogram function h (i, c) through a histogram of a mosquito image to be detected;
3.1.2) designing weighted weights of two different forms of histogram functions h (i, c) of the mosquito image to be detected under the m types of conditions to be functions g respectively according to RGB distribution characteristics of different types of mosquitoesm(i, c) and function fm(i，c)；
3.1.3) the RGB raw feature function obtained from the above function is as follows:
Figure BDA0002127343410000041
wherein, ScoremRepresenting the confidence coefficient of the mosquitoes to be detected under the m-class conditions;
3.2) simplifying the original characteristic function according to the characteristics of the specific mosquito species:
under the conditions of two types of mosquitoes, namely culex pipiens pallens and aedes albopictus, the weighting weights of two different forms of the histogram function h (i, c) are respectively as follows:
fm(i，c)＝1
Figure BDA0002127343410000042
wherein Setmc＝[1，left]∪[right，256]A Set of RGB values is obtained, and left is SetmcRight boundary of subset, right is SetmcThe left boundary of the subset;
and 3.3) obtaining the parameters of the function by using the statistical rule through the RGB distribution of the existing mosquito data.
3.3.1) setting the Score as the confidence coefficient of the mosquito to be detected under the condition only aiming at two types of mosquitoes, namely culex pipiens pallens and aedes albopictus, and setting the threshold value of the confidence coefficient of the mosquito to be detected under the condition of two types of mosquitoes, namely culex pipiens pallens and aedes albopictus as thresh; comparing the Score with the thresh, and judging the type of the mosquito to be detected: if the Score is more than thresh, the target mosquito is Aedes albopictus, otherwise, the target mosquito is Culex pipiens pallens;
3.3.2) given a Set of known RGB histogram functions of Culex pipiens and Aedes albopictus images, traversing all RGB value setsmcTwo parameters of right boundary left and left boundary right of the subset; nesting traversal of a confidence threshold thresh 'of the mosquito to be detected under the conditions of the culex pipiens pallens and the aedes albopictus in traversal to obtain a traversal value thresh' which is in a range of 0 to 1 and has a step length of 0.01;
3.3.3) sequentially substituting a given RGB histogram function and left and right boundaries left of a known mosquito image into the RGB characteristic function in traversal of thresh, taking values left 'and right' of right in the traversal, calculating to obtain a value Score 'of the RGB characteristic function, comparing the Score' with the thresh ', recording the accuracy of the classification result of the round, and obtaining the highest accuracy of the classification result by statistical ranking after the traversal of thresh is finished, wherein the corresponding parameter is left', right 'and thresh'; changing values of left and right to perform next round of traversal, comparing accuracy rates obtained in the traversal after the traversal of left and right is completely finished, and obtaining parameters with the highest accuracy rate of mosquito classification results through statistical ranking, namely the optimal parameters: left '", right'", thresh ";
3.4) combining the function g in 3.2)m(i, c) and function fm(i, c) and 3.3.3), substituting the optimal parameters left '", right'" obtained in (i, c) and 3.3.3) into the RGB raw feature function in 3.1.3) to obtain the final feature function:
Figure BDA0002127343410000051
wherein
Figure BDA0002127343410000052
And 4, calculating the confidence coefficient of the mosquitoes and classifying the target mosquitoes.
Substituting the histogram function H (i, c) of the target mosquito image obtained in the step 2 into the final characteristic function obtained in the step 3.4), calculating the confidence Score ' of the target mosquito image under the condition of only aiming at two types of mosquitoes, namely culex pipiens pallens and aedes albopictus, comparing the Score ' value with the threshold value thresh ' ″ obtained in the step 3.3.3), and judging the type of the target mosquito: if Score "> thresh'", the target mosquito is Aedes albopictus, otherwise, Culex pallidus.
The foregoing description is only an example of the present invention and is not intended to limit the invention, so that it will be apparent to those skilled in the art that various changes and modifications in form and detail may be made therein without departing from the spirit and scope of the invention.

Claims (2)

1. A mosquito identification method of a three-color RGB image is characterized by comprising the following steps:
(1) acquiring a single mosquito target image;
(2) traversing pixel points of the mosquito target image to obtain a three-color RGB image histogram of each mosquito target image and a three-color RGB image histogram function h (i, c), wherein c belongs to (R, G, B) and represents the category of three-color RGB, and i belongs to [1, 256] and represents the value of three-color RGB;
(3) designing RGB characteristic functions of different types of mosquito target images according to three-color RGB image histograms of the different types of mosquito target images:
Figure FDA0002752343560000011
wherein [ mosquito species]Is a set of mosquito species, m is a type of mosquito species set, ScoremFunction g representing confidence of the mosquito to be detected under m-type conditionsm(i, c) and function fm(i, c) represent the weighting of two different forms of histogram functions h (i, c) of the mosquito image to be detected under the m-type conditionsA weight value;
(4) simplifying the design of the function expression according to the characteristics of the specific mosquito species and obtaining the parameters of the function:
(4a) weighting weights for two different forms of an histogram function h (i, c) under two mosquito conditions of culex pipiens pallens and aedes albopictus:
fm(i，c)＝1，
Figure FDA0002752343560000012
wherein Setmc＝[1，left]∪[right，256]A Set of RGB values is obtained, and left is SetmcRight boundary of subset, right is SetmcThe left boundary of the subset;
(4b) setting the Score as the confidence coefficient of the mosquito to be detected under the condition of only aiming at two types of mosquitoes, namely culex pipiens pallens and aedes albopictus, and combining the function fm(i，c)、gm(i, c) substituting the RGB feature function in (3), thereby obtaining a simplified RGB feature function as follows:
Figure FDA0002752343560000013
(4c) setting a threshold value of confidence coefficient of the mosquitoes to be detected under the conditions of the culex pallens and the aedes albopictus as thresh, comparing the Score with the thresh, and distinguishing the types of the mosquitoes to be detected: if the Score is more than thresh, the target mosquito is Aedes albopictus, otherwise, the target mosquito is Culex pipiens pallens;
(4d) giving a group of RGB histogram functions of known culex pallens and aedes albopictus images, traversing all three parameters of left, right and thresh, sequentially substituting the RGB histogram function and the left of a given known mosquito image into the RGB characteristic function in each traversal, taking the right as a value of the traversal, calculating to obtain the value of the RGB characteristic function as Score, comparing the Score with the current value of the thresh to obtain a classification result of the round, obtaining the accuracy of the classification result under the current parameter through statistics after each traversal is finished, and obtaining the parameter with the highest accuracy of the mosquito classification result as an optimal parameter after all traversals are finished: left '", right'", thresh ";
(4e) substituting the optimal parameters left 'and right' obtained in the step (4d) into the RGB characteristic function obtained in the step (4b) to obtain a final characteristic function;
(5) calculating the confidence coefficient of the target mosquitoes and classifying the target mosquitoes:
substituting the histogram function H (i, c) of the target mosquito image obtained in the step (2) into the final characteristic function obtained in the step (4e), calculating the confidence Score ' of the target mosquito image under the condition of only aiming at two types of mosquitoes, namely culex pipiens pallens and aedes albopictus, comparing the Score ' value with the threshold value thresh ' ″ obtained in the step (4d), and judging the type of the target mosquito: if Score "> thresh'", the target mosquito is Aedes albopictus, otherwise, Culex pallidus.
2. The method for mosquito recognition of a three-color RGB image as claimed in claim 1, wherein: obtaining a single mosquito target image in the step (1), wherein the method is realized as follows:
(1a) comparing the background picture with the target image to be detected by a background subtraction method to obtain a distribution area of the mosquito target in the image; obtaining foreground image information by a background difference method, eliminating image noise by corrosion operation, and eliminating the influence of mosquito textures by expansion operation to further obtain a mosquito distribution area image;
(1b) and respectively calculating the communication area of each mosquito target on the mosquito distribution area image to obtain a single mosquito target area image.
CN201910626708.0A 2019-07-11 2019-07-11 Mosquito recognition method of three-color RGB image Active CN110516686B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910626708.0A CN110516686B (en) 2019-07-11 2019-07-11 Mosquito recognition method of three-color RGB image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910626708.0A CN110516686B (en) 2019-07-11 2019-07-11 Mosquito recognition method of three-color RGB image

Publications (2)

Publication Number Publication Date
CN110516686A CN110516686A (en) 2019-11-29
CN110516686B true CN110516686B (en) 2021-03-23

Family

ID=68622464

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910626708.0A Active CN110516686B (en) 2019-07-11 2019-07-11 Mosquito recognition method of three-color RGB image

Country Status (1)

Country Link
CN (1) CN110516686B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111160416B (en) * 2019-12-12 2023-03-21 杭州电子科技大学 Aedes mosquito identification method based on support vector machine

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102915446A (en) * 2012-09-20 2013-02-06 复旦大学 Plant disease and pest detection method based on SVM (support vector machine) learning
US9953242B1 (en) * 2015-12-21 2018-04-24 Amazon Technologies, Inc. Identifying items in images using regions-of-interest
US10007863B1 (en) * 2015-06-05 2018-06-26 Gracenote, Inc. Logo recognition in images and videos
CN108280483A (en) * 2018-01-30 2018-07-13 华南农业大学 Trypetid adult image-recognizing method based on neural network
CN108647665A (en) * 2018-05-18 2018-10-12 西安电子科技大学 Vehicle real-time detection method of taking photo by plane based on deep learning
CN109492581A (en) * 2018-11-09 2019-03-19 中国石油大学(华东) A kind of human motion recognition method based on TP-STG frame

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102915446A (en) * 2012-09-20 2013-02-06 复旦大学 Plant disease and pest detection method based on SVM (support vector machine) learning
US10007863B1 (en) * 2015-06-05 2018-06-26 Gracenote, Inc. Logo recognition in images and videos
US9953242B1 (en) * 2015-12-21 2018-04-24 Amazon Technologies, Inc. Identifying items in images using regions-of-interest
CN108280483A (en) * 2018-01-30 2018-07-13 华南农业大学 Trypetid adult image-recognizing method based on neural network
CN108647665A (en) * 2018-05-18 2018-10-12 西安电子科技大学 Vehicle real-time detection method of taking photo by plane based on deep learning
CN109492581A (en) * 2018-11-09 2019-03-19 中国石油大学(华东) A kind of human motion recognition method based on TP-STG frame

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Identifying mosquito species using smart-phone cameras;Mona Minakshi 等;《2017 European Conference on Networks and Communications (EuCNC)》;IEEE;20170615;第1-6页 *
基于ELM理论的昆虫分类;徐源浩 等;《电子科技》;20150315;第28卷(第3期);第33-37页 *

Also Published As

Publication number Publication date
CN110516686A (en) 2019-11-29

Similar Documents

Publication Publication Date Title
CN109154978B (en) System and method for detecting plant diseases
WO2022099598A1 (en) Video dynamic target detection method based on relative statistical features of image pixels
CN105404847B (en) A kind of residue real-time detection method
US8625889B2 (en) System for food recognition method using portable devices having digital cameras
CN113139521B (en) Pedestrian boundary crossing monitoring method for electric power monitoring
CN102194108B (en) Smile face expression recognition method based on clustering linear discriminant analysis of feature selection
CN108563979B (en) Method for judging rice blast disease conditions based on aerial farmland images
JP4098021B2 (en) Scene identification method, apparatus, and program
CN109063619A (en) A kind of traffic lights detection method and system based on adaptive background suppression filter and combinations of directions histogram of gradients
Masood et al. Plants disease segmentation using image processing
CN106874942B (en) Regular expression semantic-based target model rapid construction method
CN111259756A (en) Pedestrian re-identification method based on local high-frequency features and mixed metric learning
CN112257702A (en) Crop disease identification method based on incremental learning
CN105512612A (en) SVM-based image classification method for capsule endoscope
CN107784263A (en) Based on the method for improving the Plane Rotation Face datection for accelerating robust features
CN116524224A (en) Machine vision-based method and system for detecting type of cured tobacco leaves
CN110516686B (en) Mosquito recognition method of three-color RGB image
CN117765287A (en) Image target extraction method combining LWR and density clustering
Salazar-Reque et al. An image processing method to automatically identify Avocado leaf state
CN108875572B (en) Pedestrian re-identification method based on background suppression
CN110929740A (en) LGBM model-based tongue quality and tongue coating separation method
CN107341456B (en) Weather sunny and cloudy classification method based on single outdoor color image
CN112800968B (en) HOG blocking-based feature histogram fusion method for identifying identity of pigs in drinking area
Long et al. A novel unsupervised two‐stage technique in color image segmentation
CN110751023B (en) Series pedestrian detection method for video stream

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