CN115254655A - Multi-index passion fruit quality grading method based on machine vision - Google Patents
Multi-index passion fruit quality grading method based on machine vision Download PDFInfo
- Publication number
- CN115254655A CN115254655A CN202210664786.1A CN202210664786A CN115254655A CN 115254655 A CN115254655 A CN 115254655A CN 202210664786 A CN202210664786 A CN 202210664786A CN 115254655 A CN115254655 A CN 115254655A
- Authority
- CN
- China
- Prior art keywords
- passion fruit
- image
- fruit
- passion
- index
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 244000288157 Passiflora edulis Species 0.000 title claims abstract description 138
- 235000000370 Passiflora edulis Nutrition 0.000 title claims abstract description 137
- 238000000034 method Methods 0.000 title claims abstract description 27
- 235000013399 edible fruits Nutrition 0.000 claims abstract description 21
- 238000012545 processing Methods 0.000 claims abstract description 15
- 238000001514 detection method Methods 0.000 claims abstract description 10
- 238000004458 analytical method Methods 0.000 claims abstract description 6
- 238000013528 artificial neural network Methods 0.000 claims abstract description 6
- 239000002759 woven fabric Substances 0.000 claims description 12
- 238000012549 training Methods 0.000 claims description 9
- 230000000877 morphologic effect Effects 0.000 claims description 8
- 210000002615 epidermis Anatomy 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 6
- 238000012795 verification Methods 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 claims description 4
- 238000010146 3D printing Methods 0.000 claims description 3
- 239000003086 colorant Substances 0.000 claims description 3
- 230000033001 locomotion Effects 0.000 claims description 3
- 239000010902 straw Substances 0.000 claims description 3
- 238000005520 cutting process Methods 0.000 claims description 2
- 238000007639 printing Methods 0.000 claims description 2
- 238000004519 manufacturing process Methods 0.000 abstract description 7
- 238000010586 diagram Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 241000249058 Anthracothorax Species 0.000 description 1
- 235000004936 Bromus mango Nutrition 0.000 description 1
- 244000141359 Malus pumila Species 0.000 description 1
- 235000014826 Mangifera indica Nutrition 0.000 description 1
- 235000000088 Maracuja Nutrition 0.000 description 1
- 235000009184 Spondias indica Nutrition 0.000 description 1
- 235000021016 apples Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 239000002932 luster Substances 0.000 description 1
- 238000010297 mechanical methods and process Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 239000011800 void material Substances 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/04—Sorting according to size
- B07C5/10—Sorting according to size measured by light-responsive means
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/34—Sorting according to other particular properties
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C2501/00—Sorting according to a characteristic or feature of the articles or material to be sorted
- B07C2501/009—Sorting of fruit
Landscapes
- Sorting Of Articles (AREA)
Abstract
The invention discloses a multi-index passion fruit quality grading method based on machine vision, aiming at video data in a passion fruit production line, the detection of three indexes of size, maturity and whether fruit peel is shrunk is realized by combining a passion fruit shrinking condition discrimination model (namely a MobileNet V3 neural network) with traditional image processing, and the passion fruit is divided into three grades by combining a passion fruit grading standard DB 45/T2100-2019, wherein the three grades comprise the shrinking, maturing and fruit diameter range conditions. The invention solves the problems of high cost, low efficiency, high subjectivity and the like of manual classification. The invention is based on video analysis, and can be more suitable for automatic on-line operation, namely, the passion fruit can be graded nondestructively, rapidly and accurately in the conveying process of the conveyor belt. The invention integrates three indexes of the size, the maturity and the shrinking condition of the passion fruit to carry out integrated grading, and the grading precision is higher.
Description
Technical Field
The invention relates to the technical field of lossless and rapid grading of passion fruit quality, in particular to a multi-index passion fruit quality grading method based on machine vision.
Background
With the increasing public requirements on the quality of the passion fruit, the fruits with proper maturity are selected according to the transportation distance and time after being harvested, so that the optimal maturity of the fruits is guaranteed, and the competitiveness of the passion fruit market is improved.
In the existing automatic grading of the passion fruit, the size of the fruit is mostly graded only by mechanical methods such as a hole sieve and the like, indexes such as color and luster, shrinking degree and the like are ignored, but color change and shrinking condition are closely related to commercial performance of the passion fruit in the actual storage process. Although some passion fruit grading based on multiple indexes is available at present, manual evaluation is mainly used, the labor amount is large, the production rate is low, the sorting precision is easily influenced by subjective factors, and the requirements for rapidness and accuracy are difficult to realize. Therefore, the development of a multi-index comprehensive quality grading method which is automatic, lossless and suitable for the online production of passion fruit is a trend of high-quality production.
The machine vision technology can quickly extract appearance information such as target color, size and the like, is widely used for lossless and quick classification of fruits, and also utilizes machine vision to perform multi-index classification on other fruits such as mangos, apples and the like in some production lines. However, for passion fruit, the traditional mechanical or single-index grading is mainly used, and a multi-index comprehensive quality grading method combining machine vision technology and image processing is not available.
Disclosure of Invention
The invention aims to overcome the defects and shortcomings of the prior art, and provides a multi-index passion fruit quality grading method based on machine vision, aiming at video data in a passion fruit production line, the detection of three indexes of size, maturity and whether peel is shrunk is realized by combining a passion fruit shrinking condition discrimination model (namely a MobileNet V3 neural network) and traditional image processing, and the passion fruit is divided into three grades by combining a passion fruit grading standard DB 45/T2100-2019, so that the problems of high manual grading cost, low efficiency, high subjectivity and the like are solved.
In order to achieve the purpose, the technical scheme provided by the invention is as follows: a multi-index passion fruit quality grading method based on machine vision comprises the following steps:
1) Calibrating an industrial camera to obtain the actual size corresponding to a unit pixel under the height of the camera, namely the actual size of the unit pixel;
2) Placing the passion fruit on a conveyor belt, vertically placing the passion fruit on the conveyor belt by using an industrial camera, and shooting a video transmitted by the passion fruit on the conveyor belt;
3) Processing each frame image of the video in the step 2) through a background subtraction model in an OpenCV software library, identifying passion fruit and generating a binary image of the passion fruit;
4) Performing morphological closing operation processing on the binary image in the step 3) by utilizing a convolution kernel of 5 multiplied by 5 to fill a cavity in the passion fruit image, performing morphological opening operation on the image by utilizing a convolution kernel of 25 multiplied by 25 to cut off the connection between the passion fruit stem and the fruit, and preventing the interference of the fruit stem on the size detection of the passion fruit;
5) Setting the number of specific pixels on the left and right of the center of an image as a counting interval, obtaining the center x coordinate of the passion fruit based on the image processed in the step 4), starting to count the number of frames fps of the image when the center coordinate of the passion fruit enters the counting interval along with the movement of the passion fruit, taking a certain frame of image as an analysis image of three indexes of the size, the maturity and whether the peel is shrunk or not of the subsequent passion fruit, and automatically recording the sequence number of the passion fruit after the image is extracted, thereby facilitating the statistics of the subsequent result;
6) Solving a minimum external rectangle and a positive rectangle for the passion fruit outline in the image in the step 5), obtaining a longest side H2 based on the minimum external rectangle, and cutting the image based on a positive rectangle area;
7) Inputting the image cut in the step 6) into a pre-established passion fruit shrinkage condition judgment model, and detecting whether the passion fruit is shrunk or not; obtaining the actual size D of the passion fruit according to the actual size of the unit pixel provided in the step 1) and the longest side H2 of the minimum circumscribed rectangle obtained in the step 6), wherein the calculation formula is as follows: d = H2 × size; performing HSV color space conversion on the image cut in the step 6), setting a threshold based on red and yellow, reserving an area meeting the requirement of the threshold, setting other parts as backgrounds, respectively obtaining binary images of passion fruit areas under all colors, adding the binary images of red and yellow, taking the ratio of the total number of pixels of the reserved area after segmentation to the total number of pixels of the passion fruit area as a maturity judgment standard, and indicating maturity when the ratio is greater than the set value;
8) The three indexes of the size, the maturity and whether the passion fruit is shrunken are respectively scored and summed, the summed score > b is divided into a special grade, the score between [ a and b ] is divided into a first grade, and the score < a is divided into a second grade, wherein a < b.
Further, in the step 1), a white cube 3D printing piece with the side length of c is placed right below the industrial camera, a photo of the printing piece is shot, and minimum circumscribed rectangle recognition is carried out on the reference calibration object through HSV color space conversion; in order to reduce errors, the average value (H1 + W1)/2 of the sum of the length H1 and the width W1 of the minimum circumscribed rectangle is used as the side length, so that the actual size corresponding to the unit pixel, namely the actual size of the unit pixel, which is shot at the height of the camera is obtained, and the calculation is as shown in the following formula:
further, in step 2), the height of the industrial camera and the conveyor belt was set to 80cm, and the size of the photographed image was 640 × 480 pixels.
Further, in step 7), the passion fruit shrinkage condition discrimination model is pre-established, specifically, a MobileNetV3 neural network, 1000 images of the passion fruit with smooth epidermis and 500 images of the passion fruit with shrinkage of epidermis are collected under a black background and are used as a data set for training and verifying the passion fruit shrinkage condition discrimination model; performing background subtraction processing on the images in the data set based on a KNN background subtraction model in an OpenCV software library, extracting a passion fruit area, and performing image subtraction processing on the processed data set according to the following steps of (9): 1, dividing the ratio into a training set and a verification set, and inputting a passion fruit shrinkage condition judgment model for training and verification.
Further, in step 7), when the area of the reserved area is obtained, assigning 1 to the mature area in the binary image, and calculating the sum of the number of pixels in the image to obtain the total number of the pixels.
Further, in step 7), red and yellow are set at the lower threshold of each HSV channel as follows:
the H channel setting threshold is: and (3) red color: 0-woven-h-woven 10 or 156-woven-h-woven 180; yellow: 26-h woven fabric 34;
the S channel setting threshold is: red: 43-s-woven fabric 255; yellow: 43-straw bundle h-woven fabric 255;
the V channel setting threshold is: red: 46< -v < -255; yellow: 46-n-woven fabric 255;
where h denotes hue, s denotes saturation, and v denotes lightness.
Further, in the step 8), the passion fruit is divided into three grades according to the passion fruit grading standard DB 45/T2100-2019, wherein the three grades comprise whether the passion fruit is shrunken or not, whether the passion fruit is mature or not and the fruit diameter range condition.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. based on video analysis, the method can be more suitable for automatic on-line operation, namely, the passion fruit can be subjected to lossless, quick and accurate grading in the conveying process of the conveyor belt.
2. The method has the advantages that the MobileNetV3 neural network is adopted to classify whether the passion fruit peel is shrunken, network model parameters are few, the calculated amount is small, the reasoning time is short, the method is suitable for mobile embedded type edge computing equipment and the like with limited storage space and power consumption, and the passion fruit peel can be well deployed in assembly line equipment.
3. The three indexes of the passion fruit such as size, maturity and shrinkage are integrated for comprehensive grading, and grading precision is higher.
4. When the maturity is judged, the method is different from the traditional mode of counting the number of pixels of a binary image by traversing the image, the area is solved by adopting a pixel summation mode of the binary image, the detection speed is improved, and the speed of a single picture reaches about 1 ms/picture.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Fig. 2 is a flowchart of the interval range extraction image.
Fig. 3 is a schematic diagram of a binarized image obtained by a KNN-based background subtraction model.
FIG. 4 is a diagram illustrating the overall contour effect of the morphological closing operation to remove the void connection.
FIG. 5 is a diagram showing the effect of the morphological opening operation on stem removal.
Fig. 6 is a schematic diagram of an image extracted from a video for passion fruit multi-index detection.
Fig. 7 is a schematic diagram of a minimum bounding rectangle and a connected rectangle based on the passion fruit mark in the extracted image.
Fig. 8 is a schematic diagram of a passion fruit picture cut based on a connected rectangular region.
Fig. 9 is a schematic diagram of a binarized image based on the superposition of red and yellow threshold segmentation.
Fig. 10 is a diagram of the final detection effect.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the embodiments of the present invention are not limited thereto.
As shown in fig. 1, the embodiment provides a multi-index passion fruit quality grading method based on machine vision, which comprehensively grades three indexes of size, maturity and shrinkage of Tainong purple passion fruit, and the specific conditions are as follows:
1) Preparing Tainong passion fruit with smooth and shriveled epidermis, placing the Tainong passion fruit on a black background, taking a picture at the center of the visual angle of an industrial camera, collecting 1000 images of the passion fruit with the smooth and shriveled epidermis and 500 images of the passion fruit with the shriveled epidermis in total as a data set for training and verifying a pre-established passion fruit shrinkage condition discrimination model, wherein the passion fruit shrinkage condition discrimination model is a MobileNetV3 neural network specifically; further, performing background subtraction processing on the images in the data set based on a KNN background subtraction model in an OpenCV software library, and extracting a passion fruit area; the processed data set was recorded as 9:1, dividing the ratio into a training set and a verification set, and inputting a MobileNet V3 passion fruit shrinkage condition discrimination model for training and verification.
2) Calibrating a camera, placing a white cube 3D printing piece with the side length of c under an industrial camera, taking a picture of the industrial camera, and identifying the minimum external rectangle of a reference calibration object through HSV color space conversion. In order to reduce errors, the average value (H1 + W1)/2 of the sum of the length (H1) and the width (W1) of the minimum circumscribed rectangle is used as the side length, so as to obtain the actual size of the unit pixel corresponding to the shooting height of the camera, namely the actual size of the unit pixel, and the calculation is as follows:
3) In the actual detection process, the passion fruit is placed on a conveyor belt, a background subtraction model is used for identifying a binary image of the passion fruit in motion, as shown in fig. 3, the outline of a model output picture has a hole problem, the image is subjected to morphological closed operation processing through convolution checking of 5 x 5, hole filling is carried out, the whole outline is connected, and the processing result is shown in fig. 4.
4) The morphological opening operation is performed on the image by a convolution kernel of 25 × 25, the connection between the passion fruit stem and the fruit is cut off, the interference of the fruit stem on the size detection of the passion fruit is prevented, and the processing result is shown in fig. 5.
5) Setting 140 pixels on the left and right of the center of the image as counting intervals, and solving the center x coordinate of the passion fruit based on each frame of image in the video; as shown in fig. 2, as the passion fruit moves, when the central coordinate of the passion fruit enters the counting interval, the number of image frames (fps) starts to be counted, and a third frame (fps = 3) image is taken as an analysis image of three indexes of the size, the maturity and whether the peel is shrunk or not of the subsequent passion fruit, and simultaneously, after the image is extracted, the sequence number of the passion fruit is automatically recorded, so that statistics of the subsequent result is facilitated, as shown in fig. 6, the extracted image is the passion fruit at the center of the image, namely the passion fruit to be analyzed.
6) Solving the minimum circumscribed rectangle and the positive connected rectangle of the passion fruit outline at the center of the extracted image, and obtaining the longest side H2 based on the minimum circumscribed rectangle, as shown in FIG. 7; the image is cropped based on the connected rectangular area, as shown in fig. 8.
7) Detecting shrinkage indexes of passion fruits: and inputting the obtained cut image into the established passion fruit shrinking condition judgment model, so as to detect whether the passion fruit is shrunk or not.
8) And (3) detecting the size index of the passion fruit: obtaining an actual size D of the passion fruit according to the actual size of the unit pixel provided in the step 2) and the larger side length H2 of the minimum circumscribed rectangle provided in the step 6), wherein a calculation formula is as follows: d = H2 × size.
9) Detecting whether the passion fruit is mature: performing HSV color space conversion on the cut image area obtained in the step 6), setting a threshold value based on red and yellow, reserving the area meeting the threshold value requirement, and setting other parts as backgrounds so as to respectively obtain the binary images of the passion fruit areas under all colors. The red and yellow binarized images were superimposed, and as shown in fig. 9, the ratio of the total number of pixels of the area remaining after segmentation to the total number of pixels of the entire area of the passion fruit was taken as the maturity criterion, and a ratio of >60% indicates maturity.
Preferably, when the area of the reserved area is obtained, 1 is assigned to a mature area in the binary image, and the total number of pixels can be obtained by calculating the sum of the pixels in the image.
Preferably, the red and yellow thresholds are set at each HSV lane as follows:
the H channel setting threshold is: and (3) red color: 0-woven-h-woven 10 or 156-woven-h-woven 180; yellow: 26< -h < -34;
the S channel setting threshold is: red: 43-s-woven fabric 255; yellow: 43-straw bundle h-woven fabric 255;
the V channel setting threshold is: red: 46< -v < -255; yellow: 46-n-woven fabric 255;
where h denotes hue, s denotes saturation, and v denotes lightness.
10 Manual scoring and summing of three indexes of passion fruit size, maturity and whether shrinkage occurs, with the scoring criteria as follows:
TABLE 1 Manual scoring scores for the three indices
11 According to the passion fruit grading standard DB 45/T2100-2019), the passion fruit is divided into three grades, wherein the three grades comprise whether the passion fruit is shrunken or not, whether the passion fruit is ripe or not and the fruit diameter range, and the specific grading is shown in Table 2. Calculating the total score according to step 10), dividing score >6 into super grades, <5 into two grades, and the others into one grade, and the final detection effect is shown in fig. 10. The ranking according to the ranking criteria, the respective index cases and the corresponding scores included in each ranking are shown in the following table:
TABLE 2 ranking and score assignment for the cases
In conclusion, the embodiment integrates the size and the maturity of the passion fruit and whether the peel shrinks to comprehensively grade the Tainong purple fruit passion fruit, the grading precision of the special grade fruit is 94%, the grading precision of the first grade fruit is 97%, the grading precision of the second grade fruit is 100%, and the good distinguishing effect is achieved. Meanwhile, the passion fruit can be accurately counted. The method directly obtains video stream analysis through an industrial camera, the grading average precision is 97%, and the efficiency required by grading one passion fruit is 120 ms/fruit. The method can be effectively applied to automated passion fruit production line grading and is worthy of popularization.
The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.
Claims (7)
1. A multi-index passion fruit quality grading method based on machine vision is characterized by comprising the following steps:
1) Calibrating an industrial camera to obtain the actual size corresponding to a unit pixel under the height of the camera, namely the actual size of the unit pixel;
2) Placing the passion fruit on a conveyor belt, vertically placing the passion fruit on the conveyor belt by using an industrial camera, and shooting a video transmitted by the passion fruit on the conveyor belt;
3) Processing each frame image of the video in the step 2) through a background subtraction model in an OpenCV software library, identifying passion fruits and generating a binary image of the passion fruits;
4) Performing morphological closing operation processing on the binary image in the step 3) by utilizing 5 × 5 convolution kernel to fill a cavity in the passion fruit image, and performing morphological opening operation on the image by utilizing 25 × 25 convolution kernel to cut off the connection between the passion fruit stem and the fruit, so that the interference of the fruit stem on the size detection of the passion fruit is prevented;
5) Setting the number of specific pixels on the left and right of the center of an image as a counting interval, obtaining the center x coordinate of the passion fruit based on the image processed in the step 4), starting to count the number of frames fps of the image when the center coordinate of the passion fruit enters the counting interval along with the movement of the passion fruit, taking a certain frame of image as an analysis image of three indexes of the size, the maturity and whether the peel is shrunk or not of the subsequent passion fruit, and automatically recording the sequence number of the passion fruit after the image is extracted, thereby facilitating the statistics of the subsequent result;
6) Solving a minimum external rectangle and a positive rectangle for the passion fruit outline in the image in the step 5), obtaining a longest side H2 based on the minimum external rectangle, and cutting the image based on a positive rectangle area;
7) Inputting the image cut in the step 6) into a pre-established passion fruit shrinkage condition judgment model, and detecting whether the passion fruit is shrunk or not; obtaining the actual size D of the passion fruit according to the actual size of the unit pixel provided in the step 1) and the longest side H2 of the minimum circumscribed rectangle obtained in the step 6), wherein the calculation formula is as follows: d = H2 × size; performing HSV color space conversion on the image cut in the step 6), setting a threshold based on red and yellow, reserving an area meeting the requirement of the threshold, setting other parts as backgrounds, respectively obtaining binary images of passion fruit areas under all colors, adding the binary images of red and yellow, taking the ratio of the total number of pixels of the reserved area after segmentation to the total number of pixels of the passion fruit area as a maturity judgment standard, and indicating maturity when the ratio is greater than the set value;
8) The three indexes of the size, the maturity and whether the passion fruit is shrunken are respectively scored and summed, the summed score > b is divided into a special grade, the score between [ a and b ] is divided into a first grade, and the score < a is divided into a second grade, wherein a < b.
2. The machine vision-based multi-index passion fruit quality grading method according to claim 1, characterized in that: in the step 1), a white cube 3D printing piece with the side length of c is placed right below an industrial camera, a photo of the printing piece is shot, and minimum circumscribed rectangle recognition is carried out on a reference calibration object through HSV color space conversion; in order to reduce errors, the average value (H1 + W1)/2 of the sum of the length H1 and the width W1 of the minimum circumscribed rectangle is used as the side length, so that the actual size corresponding to the unit pixel, namely the actual size of the unit pixel, which is shot at the height of the camera is obtained, and the calculation is as shown in the following formula:
3. the machine vision-based multi-index passion fruit quality grading method according to claim 1, characterized in that: in step 2), the height of the industrial camera and the conveyor belt is set to 80cm, and the size of the shot image is 640 × 480 pixels.
4. The machine vision-based multi-index passion fruit quality grading method according to claim 1, characterized in that: in step 7), the passion fruit shrinkage condition discrimination model is established in advance, specifically, a mobilenetV3 neural network is used for collecting 1000 images of the passion fruit with smooth epidermis and 500 images of the passion fruit with shrunken epidermis under a black background to serve as a data set for training and verifying the passion fruit shrinkage condition discrimination model; performing background subtraction processing on the images in the data set based on a KNN background subtraction model in an OpenCV software library, extracting a passion fruit area, and performing image subtraction processing on the processed data set according to the following steps of (9): 1, dividing the ratio into a training set and a verification set, and inputting a passion fruit shrinkage condition judgment model for training and verification.
5. The machine vision-based multi-index passion fruit quality grading method according to claim 1, characterized in that: in the step 7), when the area of the reserved area is obtained, 1 is assigned to the mature area in the binary image, and the total number of pixels can be obtained by calculating the sum of the pixels in the image.
6. The machine vision-based multi-index passion fruit quality grading method according to claim 1, characterized in that: in step 7), red and yellow are set at the lower threshold of each HSV channel as follows:
the H channel setting threshold is: red: 0-woven-h-woven 10 or 156-woven-h-woven 180; yellow: 26-h woven fabric 34;
the S channel setting threshold is as follows: and (3) red color: 43-s-woven fabric 255; yellow: 43-straw bundle h-woven fabric 255;
the V channel setting threshold is: and (3) red color: 46-n-woven fabric 255; yellow: 46-n-woven fabric 255;
where h denotes hue, s denotes saturation, and v denotes lightness.
7. The machine vision-based multi-index passion fruit quality grading method according to claim 1, characterized in that: in the step 8), the passion fruit is divided into three grades according to the passion fruit grading standard DB 45/T2100-2019, wherein the three grades comprise whether the passion fruit is shrunken or not, whether the passion fruit is mature or not and the fruit diameter range.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210664786.1A CN115254655A (en) | 2022-06-14 | 2022-06-14 | Multi-index passion fruit quality grading method based on machine vision |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210664786.1A CN115254655A (en) | 2022-06-14 | 2022-06-14 | Multi-index passion fruit quality grading method based on machine vision |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115254655A true CN115254655A (en) | 2022-11-01 |
Family
ID=83759185
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210664786.1A Pending CN115254655A (en) | 2022-06-14 | 2022-06-14 | Multi-index passion fruit quality grading method based on machine vision |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115254655A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117236825A (en) * | 2023-11-15 | 2023-12-15 | 南京龟兔赛跑软件研究院有限公司 | Agricultural product supply chain transportation management system and method |
-
2022
- 2022-06-14 CN CN202210664786.1A patent/CN115254655A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117236825A (en) * | 2023-11-15 | 2023-12-15 | 南京龟兔赛跑软件研究院有限公司 | Agricultural product supply chain transportation management system and method |
| CN117236825B (en) * | 2023-11-15 | 2024-04-05 | 南京龟兔赛跑软件研究院有限公司 | Agricultural product supply chain transportation management system and method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113989279B (en) | Plastic film quality detection method based on artificial intelligence and image processing | |
| CN109961049B (en) | Cigarette brand identification method under complex scene | |
| CN111951237B (en) | Visual appearance detection method | |
| CN108562589B (en) | Method for detecting surface defects of magnetic circuit material | |
| CN109191459B (en) | Automatic identification and rating method for continuous casting billet macrostructure center segregation defect | |
| CN109772733B (en) | Defect bamboo block detecting and sorting device and method based on vision | |
| CN109978822B (en) | Banana maturity judging modeling method and judging method based on machine vision | |
| CN109145708B (en) | Pedestrian flow statistical method based on RGB and D information fusion | |
| CN107330373A (en) | A kind of parking offense monitoring system based on video | |
| CN115018838B (en) | Method for identifying pitting defects on surface of oxidized steel pipe material | |
| CN106934386B (en) | A kind of natural scene character detecting method and system based on from heuristic strategies | |
| CN108181316B (en) | Bamboo strip defect detection method based on machine vision | |
| CN106610969A (en) | Multimodal information-based video content auditing system and method | |
| CN108985170A (en) | Transmission line of electricity hanger recognition methods based on Three image difference and deep learning | |
| CN111415343A (en) | Artificial intelligence-based six-side appearance detection method for chip multilayer ceramic capacitor | |
| CN110084788B (en) | Compressed tea appearance texture recording and identifying method | |
| CN117115148B (en) | Chip surface defect intelligent identification method based on 5G technology | |
| CN109886170B (en) | Intelligent detection, identification and statistics system for oncomelania | |
| TW202005522A (en) | Fry counting system and fry counting method | |
| CN115294089A (en) | Steel surface defect detection method based on improved YOLOv5 | |
| EP3896650A1 (en) | Quality control system for series production | |
| CN115254655A (en) | Multi-index passion fruit quality grading method based on machine vision | |
| CN114985283B (en) | Tea leaf identification method based on multi-feature image identification technology | |
| CN110929787B (en) | Apple objective grading system based on image | |
| CN107145888A (en) | Video caption real time translating 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 |