CN115861308A - Disease detection method for acer truncatum - Google Patents
Disease detection method for acer truncatum Download PDFInfo
- Publication number
- CN115861308A CN115861308A CN202310145539.5A CN202310145539A CN115861308A CN 115861308 A CN115861308 A CN 115861308A CN 202310145539 A CN202310145539 A CN 202310145539A CN 115861308 A CN115861308 A CN 115861308A
- Authority
- CN
- China
- Prior art keywords
- area
- patch area
- acer truncatum
- value
- edge
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 241000219226 Acer truncatum Species 0.000 title claims abstract description 105
- 201000010099 disease Diseases 0.000 title claims abstract description 42
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 title claims abstract description 42
- 238000001514 detection method Methods 0.000 title claims abstract description 26
- 239000013598 vector Substances 0.000 claims abstract description 115
- 241000221785 Erysiphales Species 0.000 claims abstract description 75
- 238000000034 method Methods 0.000 claims abstract description 46
- 238000012545 processing Methods 0.000 claims abstract description 19
- 230000009467 reduction Effects 0.000 claims abstract description 12
- 230000002159 abnormal effect Effects 0.000 claims description 35
- 239000011159 matrix material Substances 0.000 claims description 26
- 238000003708 edge detection Methods 0.000 claims description 6
- 238000005286 illumination Methods 0.000 abstract description 8
- 230000009286 beneficial effect Effects 0.000 abstract description 6
- 230000000694 effects Effects 0.000 abstract description 5
- 230000002411 adverse Effects 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 15
- 238000004364 calculation method Methods 0.000 description 14
- 238000004458 analytical method Methods 0.000 description 7
- 238000012937 correction Methods 0.000 description 5
- 230000006378 damage Effects 0.000 description 4
- 239000003086 colorant Substances 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 241000196324 Embryophyta Species 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000011946 reduction process Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 241000238631 Hexapoda Species 0.000 description 1
- 241000607479 Yersinia pestis Species 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 239000000843 powder Substances 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of image data processing, in particular to a method for detecting diseases of acer truncatum, which comprises the following steps: obtaining each plaque area in the acer truncatum leaves to be detected, and determining the fluctuation degree and the filling degree of each plaque area according to the image characteristic information of each plaque area; correcting the weight coefficients corresponding to any two pre-acquired initial feature vectors by using the fluctuation degree and the filling degree to obtain corrected weight coefficients, and further realizing feature vector dimension reduction processing by using the corrected weight coefficients to obtain each first feature vector; and classifying the first characteristic vectors by using a nearest neighbor classifier to obtain a classification result, and further judging whether powdery mildew exists in the acer truncatum leaves to be detected. The invention increases the characteristic difference between the powdery mildew disease area and the facula area, is beneficial to reducing the adverse effect of illumination factors on the disease detection of acer truncatum, and improves the accuracy of the disease detection of acer truncatum.
Description
Technical Field
The invention relates to the technical field of image data processing, in particular to a method for detecting diseases of acer truncatum.
Background
Acer truncatum, also known as acer truncatum, is an excellent landscaping tree species. In the growth period, powdery mildew often appears in acer truncatum, which causes the back of the leaves to have large spots with unobvious edges and white mildew layers in the serious process, so that the acer truncatum leaves fall in advance. The harm to powdery mildew on acer truncatum plants is light, the plants grow badly, and the harm is serious, the plants die.
For the powdery mildew detection of acer truncatum plants in acer truncatum forests, the prior art is to convert collected leaf images from RGB space to HIS space, calculate H channel histograms by using an OSTU (maximum between-class variance threshold segmentation algorithm) threshold segmentation method to obtain local optimal thresholds, perform binarization and normalization processing on the leaf images by using the local optimal thresholds to obtain characteristic vectors, and classify the characteristic vectors by using a classifier to realize disease identification. According to the method, the characteristic vectors of the leaf images are extracted through the color information of the leaves, but under the influence of natural conditions such as illumination during image acquisition, white light spots can appear on the leaf images, and the white light spots are very similar to the image characteristics of powdery mildew, so that disease identification is performed from the color information in the leaf images independently, the condition of error identification is easy to occur, and the detection accuracy of the acer truncatum bugs is low.
Disclosure of Invention
In order to solve the technical problem of low detection accuracy of the acer truncatum buge, the invention aims to provide the acer truncatum buge disease detection method, which adopts the following technical scheme:
one embodiment of the invention provides a method for detecting diseases of acer truncatum, which comprises the following steps:
obtaining a surface image of a to-be-detected acer truncatum leaf, and performing plaque detection on the surface image to obtain each plaque area in the to-be-detected acer truncatum leaf, wherein the plaque area is an area containing powdery mildew or light spots;
obtaining an edge gradient sequence corresponding to each plaque region according to each plaque region, and determining the fluctuation degree corresponding to each plaque region according to the edge gradient sequence; determining the filling degree corresponding to each patch area according to the pixel points in each patch area;
determining an initial feature vector of each patch area according to the pixel value of each pixel point in each patch area; according to the fluctuation degree and the filling degree corresponding to each plaque area, correcting the weight coefficients corresponding to the initial feature vectors of any two pre-acquired plaque areas to obtain corrected weight coefficients;
and performing feature vector dimensionality reduction on each initial feature vector according to the corrected weight coefficient to obtain each first feature vector, classifying each first feature vector by using a nearest neighbor classifier to obtain a classification result, and judging whether powdery mildew exists in the acer truncatum leaves to be detected according to the classification result.
Further, determining the fluctuation degree corresponding to each plaque region according to the edge gradient sequence, including:
calculating a target gradient difference value corresponding to each pair of adjacent edge pixel points according to the target gradient value of each pair of adjacent edge pixel points in the edge gradient sequence corresponding to each patch area, and obtaining a differential sequence corresponding to each patch area according to the target gradient difference value;
calculating the sum of the absolute value of each difference value in the difference sequence and the value 1, taking each difference value as the numerator of the ratio, taking the sum of the absolute value of each difference value and the value 1 as the denominator of the ratio, and determining the ratio as the edge fluctuation index of the corresponding difference value;
counting the number of the edge fluctuation indexes of adjacent difference values in the difference sequence corresponding to each patch area, wherein the number of the edge fluctuation indexes of the adjacent difference values meets a preset value condition, and determining the number as the fluctuation number of the corresponding patch area, wherein the preset value condition is that one edge fluctuation index is positive and the other edge fluctuation index is negative in the edge fluctuation indexes of the adjacent difference values;
and calculating the ratio of the fluctuation quantity of each plaque area to the quantity of the difference values in the corresponding difference sequence, and determining the ratio as the fluctuation degree corresponding to the corresponding plaque area.
Further, the plaque detection is carried out on the surface image, and each plaque area in the acer truncatum leaves to be detected is obtained, and the method comprises the following steps:
clustering the surface image to obtain two clusters, calculating average pixel values corresponding to the two clusters according to pixel values of all pixel points in the two clusters, determining the pixel point in the cluster corresponding to the larger average pixel value as an abnormal pixel point, and determining the pixel point in the cluster corresponding to the smaller average pixel value as a normal pixel point;
and processing the connected domains of the abnormal pixel points in the surface image to obtain the connected domains in the surface image, and taking the closed regions formed by the connected domains as patch regions to obtain the patch regions in the acer truncatum leaves to be detected.
Further, according to the pixel points in each patch area, determining the filling degree corresponding to each patch area, including:
counting the number of normal pixels and the number of abnormal pixels in each patch area, calculating the number difference between the number of abnormal pixels and the number of normal pixels in each patch area, taking the number difference as a numerator of a ratio, taking the number of abnormal pixels in each patch area as a denominator of the ratio, and taking the ratio as a filling degree corresponding to each patch area.
Further, the formula for calculating the corrected weight coefficient is:
wherein ,
the corrected weight coefficients corresponding to the initial feature vector of the ith patch area and the initial feature vector of the jth patch area,
the fluctuation degree corresponding to the ith patch area,
the corresponding fluctuation degree of the jth plaque area,
is a pair of
The absolute value of the obtained signal is calculated,
the filling degree corresponding to the ith patch area,
for the filling degree corresponding to the jth plaque area,
is a pair of
The absolute value of the signal is calculated,
and weighting coefficients corresponding to the initial feature vector of the ith plaque area and the initial feature vector of the jth plaque area.
Further, obtaining an edge gradient sequence corresponding to each plaque region according to each plaque region includes:
performing edge detection on each patch area, obtaining each edge pixel point corresponding to each patch area, calculating the gradient value of each edge pixel point towards the direction outside the patch area, taking the gradient value towards the direction outside the patch area as a target gradient value, obtaining the target gradient value of each edge pixel point corresponding to each patch area, and taking a sequence formed by the target gradient values of each edge pixel point as an edge gradient sequence corresponding to the corresponding patch area.
Further, judging whether powdery mildew exists on the leaves of the acer truncatum bunge to be detected according to the classification result, and the method comprises the following steps:
and if the first characteristic vector with the label as the powdery mildew disease exists in the classification result, judging that the powdery mildew disease exists in the acer truncatum leaves to be detected, otherwise, judging that the powdery mildew disease does not exist in the acer truncatum leaves to be detected.
Further, determining an initial feature vector of each patch region according to the pixel value of each pixel point in each patch region, including:
constructing a minimum circumscribed rectangle of each patch area, and determining a feature matrix of each patch area according to the minimum circumscribed rectangle, wherein each numerical value in the feature matrix is a pixel value of a pixel point at a corresponding position;
and extracting a feature vector of the feature matrix according to each numerical value in the feature matrix, and determining the feature vector as an initial feature vector of the corresponding plaque area.
The invention has the following beneficial effects:
the invention provides a method for detecting diseases of acer truncatum, which is characterized in that plaque detection is carried out on surface images of acer truncatum leaves to obtain each plaque area, and each plaque area in the acer truncatum leaves is beneficial to improving the efficiency and accuracy of extracting the white powdery disease image characteristics. Because the image expression characteristics of the powdery mildew are the patch areas with unobvious edges, in order to analyze the fit degree of the image characteristics of each patch area and the image characteristics of the powdery mildew, the fluctuation degree corresponding to each patch area, namely the uniformity degree of the edge gradient of the patch area, is determined based on the edge gradient sequence corresponding to the patch area; in order to further analyze the powdery mildew degree of each patch area, the filling degree capable of representing the powdery mildew is obtained according to the distribution condition of abnormal pixel points in the patch area, and the filling degree of each patch area can reduce the adverse effect of the unstable illumination characteristic on the surface image of the acer truncatum leaf; the powdery mildew degree of each plaque area is analyzed from two angles, and the detection precision of the subsequent powdery mildew defects is improved. The powdery mildew regions and the spot regions in the images of the surfaces of the leaves of the acer truncatum have similar pixel colors, in order to enhance the differentiation characteristics between the initial characteristic vectors corresponding to the powdery mildew regions and the initial characteristic vectors corresponding to the spot regions and improve the accuracy of acer truncatum disease detection, the powdery mildew degrees of each patch region are utilized to correct the weight coefficients corresponding to the initial characteristic vectors corresponding to any two patch regions, so that the weight coefficients are fully combined with the actual powdery mildew degrees of each patch, and the association degree of the weight coefficients relative to the acer truncatum disease detection is improved. The corrected weight coefficients are utilized to realize the dimensionality reduction of the characteristic vectors, so that each first characteristic vector is obtained, the calculated amount of disease detection is effectively reduced, each first characteristic vector is classified through a nearest neighbor classifier, and a white powder disease area and a facula area can be accurately distinguished; and judging whether powdery mildew exists on the acer truncatum leaves to be detected according to the accurate classification result, so that misjudgment is reduced, and the detection precision and accuracy of the acer truncatum leaves are improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions and advantages of the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a flow chart of a disease detection method of Acer truncatum Bunge of the invention;
fig. 2 is a surface image of acer truncatum leaves in an embodiment of the invention.
Detailed Description
To further explain the technical means and effects of the present invention adopted to achieve the predetermined objects, the following detailed description of the embodiments, structures, features and effects of the technical solutions according to the present invention will be given with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" refers to not necessarily the same embodiment. Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
In the process of detecting powdery mildew of acer truncatum leaves, the acer truncatum leaves are affected by sunlight irradiation, a facula area appears in an acer truncatum leaf image, and pixel points in the facula area and pixel points in the powdery mildew area have the same color. The traditional method can realize the powdery mildew detection based on the characteristic vectors extracted from the acer truncatum leaf images, but the method can not accurately distinguish powdery mildew areas and light spot areas with similar pixel values, and the accuracy rate of the acer truncatum disease detection is seriously reduced. In order to improve the accuracy of detecting acer truncatum bugs and the difference between a stretching facula area and a powdery mildew area, the embodiment provides a method for detecting acer truncatum bugs, as shown in fig. 1, including the following steps:
s1, obtaining a surface image of the acer truncatum leaf to be detected, carrying out plaque detection on the surface image, and obtaining each plaque area in the acer truncatum leaf to be detected, wherein the steps comprise:
the method comprises the following steps of firstly, obtaining a surface image of the acer truncatum leaves to be detected.
This embodiment is when the collection waits to detect the surface image of acer truncatum leaf, uses unmanned aerial vehicle's high definition camera to carry out image acquisition through overlooking the angle to the acer truncatum leaf in the acer truncatum forest, need plan unmanned aerial vehicle's flight path in advance in the acquisition process. According to the flight route of the unmanned aerial vehicle, the surface image of the acer truncatum leaf is obtained in real time, the surface image at the current moment is used as the surface image of the acer truncatum leaf to be detected, and the surface image of the acer truncatum leaf is shown in figure 2.
So far, this embodiment carries out image acquisition through unmanned aerial vehicle to the acer truncatum leaf in the acer truncatum forest, has obtained the reference image that is used for detecting the powdery mildew disease.
And secondly, performing plaque detection on the surface image of the acer truncatum leaf to be detected to obtain each plaque area in the acer truncatum leaf to be detected.
Clustering the surface image to obtain two clusters, calculating average pixel values corresponding to the two clusters according to pixel values of all pixel points in the two clusters, determining the pixel point in the cluster corresponding to the larger average pixel value as an abnormal pixel point, and determining the pixel point in the cluster corresponding to the smaller average pixel value as a normal pixel point; and processing the connected domains of the abnormal pixel points in the surface image to obtain the connected domains in the surface image, and taking the closed regions formed by the connected domains as patch regions to obtain the patch regions in the acer truncatum leaves to be detected.
Firstly, clustering processing is carried out on the surface image to obtain average pixel values corresponding to two clusters.
In this embodiment, the surface image of the acer truncatum leaf to be detected is subjected to clustering analysis by using an FCMA (fuzzy C-means algorithm, based on a fuzzy local information C-means clustering method) clustering method with the cluster number K =2, so as to obtain two divided clusters, and the implementation process of the FCMA clustering method is the prior art, and is not described in detail here. Meanwhile, the surface image of the acer truncatum leaf to be detected is subjected to graying processing by using a weighted average method to obtain the pixel value of each pixel point in the grayscale image of the acer truncatum leaf to be detected, wherein the pixel value is the grayscale value, and the implementation process of the weighted average method is the prior art and is not elaborated herein. Mapping all the pixel points in the two clusters to the gray level image, obtaining the pixel values of all the pixel points in the two clusters, and further calculating the average pixel values corresponding to the two clusters, wherein the average pixel value is calculated to facilitate the subsequent obtaining of each abnormal pixel point in the acer truncatum leaf to be detected.
It should be noted that, when all the pixel points in the surface image of the acer truncatum leaf to be detected are divided, the FCMA clustering method can determine the cluster membership degree corresponding to each central pixel point according to the cluster membership degree corresponding to the adjacent pixel point of each central pixel point, that is, if a plurality of adjacent pixel points of a certain central pixel point are abnormal pixel points, the central pixel point should also be an abnormal pixel point. Compared with other existing clustering methods, the FCMA clustering method can retain local information, is beneficial to improving the dividing accuracy of two divided clusters, and is more suitable for clustering processing of surface images of acer truncatum leaves.
And then, obtaining each abnormal pixel point in the acer truncatum leaves to be detected according to the average pixel values corresponding to the two clusters.
It should be noted that the colors of the normal areas of the acer truncatum leaves in the same growth cycle are the same, and the powdery mildew of the acer truncatum leaves is characterized in that a large number of powdery mildew pixel points exist on the acer truncatum leaves, the powdery mildew pixel points are greatly different from the pixel points in the normal areas, and in order to analyze whether the powdery mildew pixel points exist in the acer truncatum leaves to be detected, the pixel points with similar colors are clustered and divided into normal pixel points and abnormal pixel points by means of pixel point clustering on a surface image.
In this embodiment, average pixel values corresponding to two clusters are compared, in combination with the image characteristics of the powdery mildew, a pixel point in a cluster corresponding to a larger average pixel value is determined as an abnormal pixel point, a pixel point in a cluster corresponding to a smaller average pixel value is determined as a normal pixel point, the abnormal pixel point refers to a powdery mildew pixel point or a spot pixel point, the pixel values of the powdery mildew pixel point and the spot pixel point are both larger, and the normal pixel point refers to a pixel point in a normal area which is not affected by the powdery mildew and the illumination.
And finally, obtaining each patch area in the acer truncatum leaves to be detected according to each abnormal pixel point in the acer truncatum leaves to be detected.
In this embodiment, connected domain processing is performed on each abnormal pixel point in the acer truncatum leaf to be detected to obtain each connected domain in the surface image, a closed area formed by the connected domains is used as a patch area, each patch area in the acer truncatum leaf to be detected can be obtained, the patch area is an area containing powdery mildew or light spots, and the patch area refers to a multi-connected area which is not all of abnormal pixel points but part of which is surrounded by a closed curve. For the ith patch area in the acer truncatum leaf to be detected, recording the number of all abnormal pixel points in the ith patch area as
Recording the number of all normal pixel points in the ith patch area as
The implementation process of connected domain processing is prior art and is not within the scope of the present invention, and is not described in detail herein.
S2, obtaining an edge gradient sequence corresponding to each plaque region according to each plaque region, and determining the fluctuation degree corresponding to each plaque region according to the edge gradient sequence; determining the filling degree corresponding to each plaque area according to the pixel points in each plaque area, wherein the steps comprise:
and step one, obtaining an edge gradient sequence corresponding to each plaque area according to each plaque area.
Performing edge detection on each patch area, obtaining each edge pixel point corresponding to each patch area, calculating the gradient value of each edge pixel point towards the direction outside the patch area, taking the gradient value towards the direction outside the patch area as a target gradient value, obtaining the target gradient value of each edge pixel point corresponding to each patch area, and taking a sequence formed by the target gradient values of each edge pixel point as an edge gradient sequence corresponding to the corresponding patch area.
In this implementation, sobel edge detection operators are used to perform edge detection on each patch area to obtain each edge pixel point corresponding to each patch area, and the implementation process of Sobel edge detection operators is the prior art and will not be elaborated here. Because the edge pixel points are similar to the pixel points in the patch area, the gradient value of the edge pixel points towards the inner direction of the patch area is calculated, the gradient value of the edge pixel points is low, no reference value exists, the gradient value of each edge pixel point towards the outer direction of the patch area is extracted according to the clockwise direction, the gradient value towards the outer direction of the patch area is used as a target gradient value, the target gradient value of each edge pixel point corresponding to each patch area is obtained, and the target gradient value of the jth edge pixel point corresponding to each patch area is recorded as a target gradient value
. The target gradient values of the edge pixel points corresponding to each patch area can form an edge gradient sequence, so that the edge gradient sequence corresponding to each patch area is obtained, and the edge gradient sequence is beneficial to subsequent analysis of the fluctuation degree corresponding to each patch area.
It should be noted that, the calculation process of the gradient values of the edge pixels is the prior art, and is not within the protection scope of the present invention, and is not described in detail herein, the clockwise extraction of the gradient values is adopted here to facilitate the subsequent analysis and calculation, and of course, the implementer may also adopt the counterclockwise extraction of the gradient values.
And secondly, determining the fluctuation degree corresponding to each plaque area according to the edge gradient sequence.
It should be noted that powdery mildew can cause the acer truncatum leaves to have patches with unobvious edges, so based on the image characteristics of the powdery mildew defect, if a certain patch area is a powdery mildew area, the gradient value fluctuation of the edges of the patch area is relatively large; based on the image characteristics of the spot area, if a certain patch area is the spot area, the fluctuation of the gradient value of the edge of the patch area is small, and the gradient value distribution is uniform. Therefore, the fluctuation degree corresponding to each plaque area is analyzed through the uniformity degree of the edge gradient sequence corresponding to each plaque area, wherein the fluctuation degree can represent the powdery mildew degree of the plaque area, and the method comprises the following steps:
firstly, calculating a target gradient difference value corresponding to each pair of adjacent edge pixel points according to the target gradient value of each pair of adjacent edge pixel points in the edge gradient sequence corresponding to each patch area, and obtaining a difference sequence corresponding to each patch area according to the target gradient difference value.
In this embodiment, the calculation formula of the target gradient difference value corresponding to each pair of adjacent edge pixel points may be:
wherein ,
the target gradient difference value corresponding to the nth edge pixel point and the (n + 1) th edge pixel point in the edge gradient sequence corresponding to each plaque area is defined as a pair of adjacent edge pixel points,
for the target gradient value of the (n + 1) th edge pixel point in the edge gradient sequence corresponding to each plaque region,
and obtaining the target gradient value of the nth edge pixel point in the edge gradient sequence corresponding to each plaque area.
In the calculation formula of the target gradient difference value,
can represent the difference condition of the target gradient values corresponding to the adjacent edge pixel points, and each target gradient difference value corresponding to each plaque area
The composed sequence can be regarded as a differential sequence, and each plaque area has its corresponding differential sequence. The value in the difference sequence is a difference value, the target gradient difference value is a difference value, and the difference value can represent the target gradient value change condition of the front and rear adjacent edge pixel points.
Then, calculating the sum of the absolute value of each difference value in the difference sequence and the value 1, taking each difference value as the numerator of the ratio, taking the sum of the absolute value of each difference value and the value 1 as the denominator of the ratio, and determining the ratio as the edge fluctuation index of the corresponding difference value.
In this embodiment, when the number of times of change of the positive and negative numbers in the difference sequence is relatively large, it indicates that the fluctuation of the edge of the plaque area corresponding to the difference sequence is relatively large, and the plaque area corresponding to the difference sequence has a relatively high possibility of having the powdery mildew defect. For the positive-negative relation of each difference value in the difference sequence, the positive-negative relation is obtained by the 1 norm of each difference value, that is, the difference value is divided by the sum of the absolute value of the difference value and the value 1, where the value 1 is to avoid the special case that the fraction denominator is 1, the characterization form of the edge fluctuation index may be "+" or "-", and the calculation formula of the edge fluctuation index may be:
wherein ,
for the m-th edge fluctuation index in the differential sequence corresponding to each plaque region,
for the mth difference value in the difference sequence corresponding to each plaque region,
to pair
And (6) calculating an absolute value.
Then, the number of the edge fluctuation indexes of the adjacent difference values in the difference sequence corresponding to each plaque area, which meet the preset numerical value condition, is counted, and the number is determined as the fluctuation number of the corresponding plaque area.
In this embodiment, the number of adjacent edge fluctuation indexes meeting a preset numerical condition is counted according to the edge fluctuation indexes of adjacent difference values in the difference sequence corresponding to each patch area, where the preset numerical condition is that one edge fluctuation index is positive and the other edge fluctuation index is negative in the edge fluctuation indexes of the adjacent difference values. It should be noted that, when a pair of adjacent edge fluctuation indexes satisfies a preset numerical condition, the positive-negative relationship of the pair of adjacent edge fluctuation indexes is not divided, that is, a first edge fluctuation index of the pair of adjacent edge fluctuation indexes may be a positive number, a second edge fluctuation index of the pair of adjacent edge fluctuation indexes may be a negative number, and similarly, the first edge fluctuation index of the pair of adjacent edge fluctuation indexes may be a negative number, and the second edge fluctuation index of the pair of adjacent edge fluctuation indexes may be a positive number.
For example, a dimension is established as
The sliding window is made to slide in the difference sequence corresponding to each patch area, the sliding step length is 1, based on the edge fluctuation index of each difference value in the difference sequence, every time when the edge fluctuation index of two negative and positive difference values appears in the sliding window, the accumulator is added with 1 until the sliding window is finished, the accumulator value of the difference sequence corresponding to each patch area can be obtained, and the accumulator value is used as the fluctuation quantity of the corresponding patch area.
And finally, calculating the ratio of the fluctuation quantity of each plaque area to the quantity of the difference values in the corresponding difference sequence, and determining the ratio as the fluctuation degree corresponding to the corresponding plaque area.
In this embodiment, only one fluctuation state exists between every two adjacent difference values in the difference sequence, so that the fluctuation number of each patch region is inevitably smaller than the number of the difference values in the difference sequence corresponding to the corresponding patch region, and the fluctuation degree of the current patch region can be represented by the ratio of the fluctuation number to the number of the difference values, and the calculation formula may be:
wherein ,
as the fluctuation degree of the a-th patch area,
the fluctuation amount of the a-th patch area,
the number of the difference values in the difference sequence corresponding to the a-th plaque area.
It should be noted that the fluctuation degree of the patch area can represent the fluctuation condition of the edge pixel points of the patch area, and in the subsequent powdery mildew detection analysis process, the fluctuation condition of the edge pixel points of each patch area can measure whether each patch area is caused by the powdery mildew or the illumination spots.
And thirdly, determining the filling degree corresponding to each patch area according to the pixel points in each patch area.
Counting the number of normal pixels and the number of abnormal pixels in each patch area, calculating the difference between the number of abnormal pixels and the number of normal pixels in each patch area, taking the difference as the numerator of the ratio, taking the number of abnormal pixels in each patch area as the denominator of the ratio, and taking the ratio as the corresponding filling degree of each patch area.
First, when a leaf damage caused by an insect pest occurs in a acer truncatum leaf, the leaf damage is influenced by unstable illumination characteristics, and when illumination penetrates through a damaged leaf area, the leaf below the damaged leaf area may also have a light spot influence with strong edge fluctuation. Therefore, while determining the fluctuation degree of each patch area, the filling degree of the abnormal pixel points in each patch area also needs to be considered, and from the two angles of the fluctuation degree and the filling degree, the formation reason of each patch area is analyzed to obtain the powdery mildew degree of each patch area.
In this embodiment, the filling degree of the patch area is determined by the number of normal pixels and the number of abnormal pixels in the patch area. When the number of normal pixel points inside a closed area formed by a patch area is more, the more uneven the distribution of abnormal pixel points in the patch area is, on the basis of analyzing powdery mildew by fluctuation degree, the distribution uniformity degree of the abnormal pixel points inside the patch area needs to be further analyzed, and the calculation formula of the filling degree corresponding to each patch area can be as follows:
wherein ,
the filling degree corresponding to the ith patch area,
the number of abnormal pixel points in the ith patch area,
the number of normal pixels in the ith patch area,
the number difference between the number of the abnormal pixel points and the number of the normal pixel points in the ith patch area is obtained.
In the calculation formula of the filling degree, the number of normal pixel points in the ith patch area
The larger the difference value between the number of abnormal pixels and the number of normal pixels in the ith patch area
The smaller the size, the corresponding filling of the ith plaque areaDegree 
The smaller will be.
Therefore, the fluctuation degree corresponding to each patch area can be obtained through the analysis of the fluctuation condition of the edge pixel points in each patch area, and the filling degree corresponding to each patch area can be obtained through the analysis of the filling degree of the abnormal pixel points in each patch area. The higher the fluctuation degree and the filling degree corresponding to a certain patch area are, the more the patch area conforms to the image characteristics of the powdery mildew, namely the fluctuation degree and the filling degree can be used as the index of the powdery mildew degree of the acer truncatum leaves, and the index can be used for correcting the subsequently extracted feature vector, so that the distance between the leaf area with the facula and the leaf area with the powdery mildew is increased, and the accuracy of detecting the acer truncatum bugs is improved.
S3, determining an initial feature vector of each patch area according to the pixel value of each pixel point in each patch area; according to the corresponding fluctuation degree and filling degree of each patch area, correcting the weight coefficients corresponding to the initial feature vectors of any two pre-acquired patch areas to obtain corrected weight coefficients, wherein the steps comprise:
the method comprises the following steps that firstly, according to the pixel value of each pixel point in each patch area, an initial feature vector of each patch area is determined.
Constructing a minimum circumscribed rectangle of each patch area, and determining a feature matrix of each patch area according to the minimum circumscribed rectangle, wherein each numerical value in the feature matrix is a pixel value of a pixel point at a corresponding position; and extracting a feature vector of the feature matrix according to each numerical value in the feature matrix, and determining the feature vector as an initial feature vector of the corresponding plaque area.
In the present embodiment, a minimum bounding rectangle for each patch region is constructed, and the process of constructing the minimum bounding rectangle is prior art and will not be described in detail here. Because the feature matrix is a square matrix, the feature matrix corresponding to each patch area is convenient to determine subsequently, the width value of the minimum circumscribed rectangle is expanded to the length value based on the length value of the rectangle, and the widths of the two sides are expanded simultaneously during expansion to obtain the circumscribed square of each patch area. And obtaining a characteristic matrix according to the external square in an analogy manner, wherein each patch area has a corresponding characteristic matrix, each numerical value in each characteristic matrix is the pixel value of a pixel point at a corresponding position, and if the numerical value which does not belong to the corresponding patch area exists in the characteristic matrix, the numerical value is assigned to be 0. By means of the feature matrixes corresponding to the patch areas, feature vectors corresponding to the feature matrixes can be obtained by utilizing a calculation method of feature vectors of the existing matrixes, the number of the feature vectors of each feature matrix can be 1 or 2, and one of the feature vectors is randomly selected to serve as an initial feature vector of the corresponding feature matrix. It should be noted that the eigenvectors of the feature matrix are determined according to the dimension of the matrix, and the grayscale image matrix is a two-dimensional matrix, and since there is only one real solution, the number of eigenvectors of each feature matrix may be 1 or 2.
Thus, initial feature vectors of the respective patch regions are obtained, and the initial feature vectors can represent image feature information of the patch regions.
And secondly, correcting the weight coefficients corresponding to the initial feature vectors of any two patch areas according to the corresponding fluctuation degree and filling degree of each patch area to obtain the corrected weight coefficients.
First, in the LPP (local Preserving projection) projection algorithm, the objective function analyzes the weight coefficient between one feature vector and another feature vector by the distance between the two feature vectors, and the two feature vectors are not identical but very similar. In order to facilitate the subsequent correction of the weighting coefficients based on the powdery mildew degree corresponding to each patch area, the distance between the feature vector corresponding to each patch area and the feature vector corresponding to each spot area is stretched, and the weighting coefficients corresponding to the initial feature vectors of any two patch areas need to be obtained, and the calculation formula can be as follows:
wherein ,
the weighting coefficients corresponding to the initial feature vector of the ith patch area and the initial feature vector of the jth patch area, e is a natural constant,
is the initial feature vector of the ith patch region,
is the initial feature vector of the jth patch area, t is a hyperparameter, t =5 in this embodiment,
is composed of
And with
The distance between the two or more of the two or more,
is based on a natural constant e
Negative correlation index of (c).
It should be noted that, the calculation formula of the weight coefficients corresponding to the two initial feature vectors is an existing formula, the hyperparameter t is set artificially, and represents the bandwidth of the gaussian kernel function, and when the weight coefficients corresponding to the two initial feature vectors are 1, it indicates that the two initial feature vectors are the same feature vector. For any two initial feature vectors in the feature space
And
the projection may be achieved by one initial feature vector between two initial feature vectors.
And then, according to the corresponding fluctuation degree and filling degree of each patch area, correcting the weight coefficients corresponding to the initial feature vectors of any two pre-acquired patch areas to obtain the corrected weight coefficients.
In this embodiment, the fluctuation degree and the filling degree can be used as correction factors in the feature dimension reduction process. In the process of performing feature dimension reduction processing on each initial feature vector by using an LPP projection algorithm, when the distance between the initial feature vectors of the powdery mildew area and the spot area is calculated, the powdery mildew area and the spot area cannot be effectively distinguished due to the fact that any two initial feature vectors are very similar, and the powdery mildew area and the spot area are the spot area. For distinguishing the powdery mildew area and the spot area, the weight coefficient of the distance between the powdery mildew area and the spot area in the projection process needs to be increased, that is, in order to stretch the distance between the powdery mildew area and the initial feature vector of the spot area, and to retain the local image information of the spot area, the weight coefficient needs to be corrected by two correction factors, namely the fluctuation degree and the filling degree, to obtain a corrected weight coefficient, and the calculation formula can be as follows:
wherein ,
the corrected weight coefficients corresponding to the initial feature vector of the ith patch area and the initial feature vector of the jth patch area,
the fluctuation degree corresponding to the ith patch area,
the corresponding fluctuation degree of the jth plaque area,
is a pair of
The absolute value of the signal is calculated,
the filling degree corresponding to the ith plaque area,
for the filling degree corresponding to the jth plaque area,
is a pair of
The absolute value of the obtained signal is calculated,
and weighting coefficients corresponding to the initial feature vector of the ith patch area and the initial feature vector of the jth patch area.
In the calculation formula of the corrected weight coefficient,
the difference of the fluctuation degrees corresponding to the two patch areas can be represented, and when the fluctuation degree difference corresponding to the two patch areas is larger, the larger the dimension reduction weight coefficient corresponding to the initial feature vectors of the two patch areas needs to be increased, so that the two initial feature vectors realize differentiation through the white powdery mildew image features of the patch areas;
the difference of the filling degree of the two plaque areas can be represented,
influence on initial feature vector
The influence on the initial characteristic vector is the same, the filling degree is further corrected on the basis of the fluctuation degree, so that the accuracy of the corrected weight coefficient is improved, and the unstable characteristic caused by illumination is relievedAdverse effects on the extracted powdery mildew image characteristics; the 1 in the calculation formula is to avoid a special case where the weight coefficient after correction is 0.
And S4, performing feature vector dimensionality reduction on each initial feature vector according to the corrected weight coefficient to obtain each first feature vector, classifying each first feature vector by using a nearest neighbor classifier to obtain a classification result, and judging whether powdery mildew exists on the leaves of the acer truncatum to be detected according to the classification result.
And step one, performing feature vector dimensionality reduction processing on each initial feature vector according to the corrected weight coefficient to obtain each first feature vector, and classifying each first feature vector by using a nearest neighbor classifier to obtain a classification result.
In this embodiment, first, the corrected weight coefficient is applied to the LPP projection algorithm to obtain a new LPP projection algorithm, and the feature vector dimension reduction processing is performed on the initial feature vector corresponding to each plaque region by using the new LPP projection algorithm to obtain a feature vector after the dimension reduction processing, and the feature vector after the dimension reduction processing is used as the first feature vector. The implementation process of the LPP projection algorithm is prior art and is not within the scope of the present invention, and is not described in detail herein. Then, classifying each first feature vector by using a nearest neighbor classifier, labeling a corresponding label for each first feature vector, and obtaining a final classification result of the acer truncatum leaves to be detected. The process of implementing the classification process by the nearest neighbor classifier is prior art and is not within the scope of the present invention, and is not described in detail herein.
It should be noted that any two initial feature vectors have their corresponding corrected weight coefficients, and the LPP projection algorithm can ensure that adjacent samples in each initial feature vector are still adjacent in each first feature vector, that is, the feature vectors after dimensionality reduction retain original neighboring information, and the retention of the original neighboring information is beneficial for a nearest neighbor classifier to accurately classify the leaf disease information of each first feature vector, so as to divide the leaf disease information into a first feature vector with powdery mildew and a first feature vector without powdery mildew, that is, a plaque area with powdery mildew and a plaque area without powdery mildew.
And secondly, judging whether powdery mildew diseases exist in the leaves of the acer truncatum bunge to be detected according to classification results.
In this embodiment, if the first feature vector with the label of powdery mildew exists in the final classification result, it is determined that powdery mildew exists on the acer truncatum leaves to be detected, otherwise, it is determined that powdery mildew does not exist on the acer truncatum leaves to be detected.
So far, this embodiment has realized acer truncatum disease detection through carrying out image data processing to acer truncatum leaf surface image.
The invention provides a method for detecting diseases of acer truncatum, which comprises the steps of carrying out image data processing and analysis on surface images of acer truncatum leaves to obtain fluctuation degrees and filling degrees of weight coefficients used for correcting a local maintenance projection algorithm, stretching powdery mildew areas and spot areas in a dimension reduction process based on two correction factors, namely changing initial characteristic vectors of corresponding patch areas according to the disease degrees of the patch areas, facilitating the updating of the initial characteristic vectors of the patch areas based on the disease degrees of the corresponding patch areas, and further improving the detection accuracy of the acer truncatum diseases.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present application, and not for limiting the same; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; the modifications or substitutions do not make the essence of the corresponding technical solutions deviate from the technical solutions of the embodiments of the present application, and are included in the protection scope of the present application.
Claims (8)
1. A method for detecting diseases of acer truncatum, which is characterized by comprising the following steps:
obtaining a surface image of a to-be-detected acer truncatum leaf, and performing plaque detection on the surface image to obtain each plaque area in the to-be-detected acer truncatum leaf, wherein the plaque area is an area containing powdery mildew or light spots;
obtaining an edge gradient sequence corresponding to each plaque region according to each plaque region, and determining the fluctuation degree corresponding to each plaque region according to the edge gradient sequence; determining the filling degree corresponding to each patch area according to the pixel points in each patch area;
determining an initial feature vector of each patch area according to the pixel value of each pixel point in each patch area; according to the fluctuation degree and the filling degree corresponding to each plaque area, correcting the weight coefficients corresponding to the initial feature vectors of any two pre-acquired plaque areas to obtain corrected weight coefficients;
and performing feature vector dimensionality reduction on each initial feature vector according to the corrected weight coefficient to obtain each first feature vector, classifying each first feature vector by using a nearest neighbor classifier to obtain a classification result, and judging whether powdery mildew exists in the acer truncatum leaves to be detected according to the classification result.
2. The method for detecting diseases of acer truncatum, according to claim 1, wherein determining the fluctuation degree corresponding to each plaque area according to the edge gradient sequence comprises:
calculating a target gradient difference value corresponding to each pair of adjacent edge pixel points according to the target gradient value of each pair of adjacent edge pixel points in the edge gradient sequence corresponding to each patch area, and obtaining a differential sequence corresponding to each patch area according to the target gradient difference value;
calculating the sum of the absolute value of each difference value in the difference sequence and the value 1, taking each difference value as the numerator of the ratio, taking the sum of the absolute value of each difference value and the value 1 as the denominator of the ratio, and determining the ratio as the edge fluctuation index of the corresponding difference value;
counting the number of the edge fluctuation indexes of adjacent difference values in the difference sequence corresponding to each patch area, wherein the number of the edge fluctuation indexes of the adjacent difference values meets a preset value condition, and determining the number as the fluctuation number of the corresponding patch area, wherein the preset value condition is that one edge fluctuation index is positive and the other edge fluctuation index is negative in the edge fluctuation indexes of the adjacent difference values;
and calculating the ratio of the fluctuation quantity of each plaque area to the quantity of the difference values in the corresponding difference sequence, and determining the ratio as the fluctuation degree corresponding to the corresponding plaque area.
3. The method for detecting diseases of acer truncatum, according to claim 1, wherein the step of performing plaque detection on the surface image to obtain each plaque area in acer truncatum leaves to be detected comprises the following steps:
clustering the surface image to obtain two clusters, calculating average pixel values corresponding to the two clusters according to pixel values of all pixel points in the two clusters, determining the pixel point in the cluster corresponding to the larger average pixel value as an abnormal pixel point, and determining the pixel point in the cluster corresponding to the smaller average pixel value as a normal pixel point;
and processing the connected domains of the abnormal pixel points in the surface image to obtain the connected domains in the surface image, and taking the closed regions formed by the connected domains as patch regions to obtain the patch regions in the acer truncatum leaves to be detected.
4. The method for detecting diseases of acer truncatum, according to claim 3, wherein determining the filling degree corresponding to each patch area according to the pixel points in each patch area comprises:
counting the number of normal pixels and the number of abnormal pixels in each patch area, calculating the number difference between the number of abnormal pixels and the number of normal pixels in each patch area, taking the number difference as a numerator of a ratio, taking the number of abnormal pixels in each patch area as a denominator of the ratio, and taking the ratio as a filling degree corresponding to each patch area.
5. The method for detecting diseases of acer truncatum as claimed in claim 1, wherein the formula for calculating the corrected weight coefficient is as follows:
wherein ,
the corrected weighting coefficients corresponding to the initial feature vector of the ith patch area and the initial feature vector of the jth patch area are determined>
Based on the degree of fluctuation associated with the ith patch area>
Based on the degree of fluctuation corresponding to the jth patch area>
Is a pair>
Evaluating an absolute value>
Based on the filling degree corresponding to the ith plaque area>
For the filling degree corresponding to the jth patch area, <' >>
Is a pair>
Evaluating an absolute value>
And weighting coefficients corresponding to the initial feature vector of the ith patch area and the initial feature vector of the jth patch area.
6. The method for detecting diseases of acer truncatum, according to claim 1, wherein obtaining the edge gradient sequence corresponding to each plaque area according to each plaque area comprises:
performing edge detection on each patch area, obtaining each edge pixel point corresponding to each patch area, calculating the gradient value of each edge pixel point towards the direction outside the patch area, taking the gradient value towards the direction outside the patch area as a target gradient value, obtaining the target gradient value of each edge pixel point corresponding to each patch area, and taking a sequence formed by the target gradient values of each edge pixel point as an edge gradient sequence corresponding to the corresponding patch area.
7. The method for detecting diseases of acer truncatum as claimed in claim 1, wherein judging whether powdery mildew exists on leaves of acer truncatum to be detected according to classification results comprises:
and if the first characteristic vector with the label as the powdery mildew disease exists in the classification result, judging that the powdery mildew disease exists in the acer truncatum leaves to be detected, otherwise, judging that the powdery mildew disease does not exist in the acer truncatum leaves to be detected.
8. The method for detecting diseases of acer truncatum, according to claim 1, wherein determining the initial feature vector of each patch area according to the pixel value of each pixel point in each patch area comprises:
constructing a minimum circumscribed rectangle of each patch area, and determining a feature matrix of each patch area according to the minimum circumscribed rectangle, wherein each numerical value in the feature matrix is a pixel value of a pixel point at a corresponding position;
and extracting a feature vector of the feature matrix according to each numerical value in the feature matrix, and determining the feature vector as an initial feature vector of the corresponding plaque area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310145539.5A CN115861308B (en) | 2023-02-22 | 2023-02-22 | Acer truncatum disease detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310145539.5A CN115861308B (en) | 2023-02-22 | 2023-02-22 | Acer truncatum disease detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115861308A true CN115861308A (en) | 2023-03-28 |
| CN115861308B CN115861308B (en) | 2023-05-12 |
Family
ID=85658591
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310145539.5A Active CN115861308B (en) | 2023-02-22 | 2023-02-22 | Acer truncatum disease detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115861308B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117437248A (en) * | 2023-12-19 | 2024-01-23 | 中国海洋大学 | Disease area segmentation method for blueberry quality detection |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104398271A (en) * | 2014-11-14 | 2015-03-11 | 西安交通大学 | Method using three-dimensional mechanics and tissue specific imaging of blood vessels and plaques for detection |
| CN104598908A (en) * | 2014-09-26 | 2015-05-06 | 浙江理工大学 | Method for recognizing diseases of crop leaves |
| CN106778786A (en) * | 2016-12-29 | 2017-05-31 | 西京学院 | Apple disease recognition methods based on log-spectral domain laminated gradient direction histogram |
| CN108764275A (en) * | 2018-04-10 | 2018-11-06 | 甘肃农业大学 | The recognition methods of blade disease and system |
| WO2020206666A1 (en) * | 2019-04-12 | 2020-10-15 | 深圳市汇顶科技股份有限公司 | Depth estimation method and apparatus employing speckle image and face recognition system |
| WO2021179590A1 (en) * | 2020-03-10 | 2021-09-16 | 北京迈格威科技有限公司 | Disparity map processing method and apparatus, computer device and storage medium |
-
2023
- 2023-02-22 CN CN202310145539.5A patent/CN115861308B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104598908A (en) * | 2014-09-26 | 2015-05-06 | 浙江理工大学 | Method for recognizing diseases of crop leaves |
| CN104398271A (en) * | 2014-11-14 | 2015-03-11 | 西安交通大学 | Method using three-dimensional mechanics and tissue specific imaging of blood vessels and plaques for detection |
| CN106778786A (en) * | 2016-12-29 | 2017-05-31 | 西京学院 | Apple disease recognition methods based on log-spectral domain laminated gradient direction histogram |
| CN108764275A (en) * | 2018-04-10 | 2018-11-06 | 甘肃农业大学 | The recognition methods of blade disease and system |
| WO2020206666A1 (en) * | 2019-04-12 | 2020-10-15 | 深圳市汇顶科技股份有限公司 | Depth estimation method and apparatus employing speckle image and face recognition system |
| WO2021179590A1 (en) * | 2020-03-10 | 2021-09-16 | 北京迈格威科技有限公司 | Disparity map processing method and apparatus, computer device and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| JAEMYUNG SHIN等: ""Effect of directional augmentation using supervised machine learning technologies: A case study of strawberry powdery mildew detection"" * |
| 焦翠玲等: ""基于超像素的农作物病害图像分割方法研究"" * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117437248A (en) * | 2023-12-19 | 2024-01-23 | 中国海洋大学 | Disease area segmentation method for blueberry quality detection |
| CN117437248B (en) * | 2023-12-19 | 2024-03-08 | 中国海洋大学 | Disease area segmentation method for blueberry quality detection |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115861308B (en) | 2023-05-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115018828B (en) | Defect detection method for electronic component | |
| CN107610114B (en) | optical satellite remote sensing image cloud and snow fog detection method based on support vector machine | |
| Yu et al. | Automatic grading of apples based on multi-features and weighted K-means clustering algorithm | |
| US8340420B2 (en) | Method for recognizing objects in images | |
| Sunny et al. | An efficient citrus canker detection method based on contrast limited adaptive histogram equalization enhancement | |
| CN106845497B (en) | Corn early-stage image drought identification method based on multi-feature fusion | |
| CN111882586B (en) | Multi-actor target tracking method oriented to theater environment | |
| CN112464942B (en) | Computer vision-based overlapped tobacco leaf intelligent grading method | |
| CN114332650B (en) | Remote sensing image road identification method and system | |
| CN103415871B (en) | Irradiation field identifies | |
| CN107944354B (en) | Vehicle detection method based on deep learning | |
| CN109685045A (en) | A kind of Moving Targets Based on Video Streams tracking and system | |
| CN115410096B (en) | Satellite remote sensing image multi-scale fusion change detection method, medium and electronic device | |
| CN115861308A (en) | Disease detection method for acer truncatum | |
| CN115908371A (en) | Plant leaf disease and insect pest degree detection method based on optimized segmentation | |
| CN115690086A (en) | Object-based high-resolution remote sensing image change detection method and system | |
| Quispe et al. | Automatic building change detection on aerial images using convolutional neural networks and handcrafted features | |
| CN113344047A (en) | Platen state identification method based on improved K-means algorithm | |
| CN116206208B (en) | Forestry plant diseases and insect pests rapid analysis system based on artificial intelligence | |
| CN106446832B (en) | Video-based pedestrian real-time detection method | |
| CN111144280A (en) | Monitoring video leaf occlusion detection method | |
| Yang et al. | License plate detection based on sparse auto-encoder | |
| CN107341456B (en) | Weather sunny and cloudy classification method based on single outdoor color image | |
| Raj et al. | Automatic defect identification and grading system for ‘Jonagold’apples | |
| CN115862003A (en) | Lightweight YOLOv 5-based in-vivo apple target detection and classification method |
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 |