System and method for detecting impurities in bottled linseed oil
Through machine vision technology, flaxseed oil is photographed and imaged in multiple directions, detecting and identifying its impurities, solving the problem of impurities detection in flaxseed oil and ensuring the safety of the oil.
Patent Information
- Application Number
- CN202411523766.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-30
- Publication Date
- 2024-12-27
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Flaxseed oil may have impurities during the refining process, affecting its quality and safety. It is difficult for the existing technology to effectively detect and identify impurities, resulting in health problems such as food poisoning.
A bottled flaxseed oil impurity detection system is adopted, including an image acquisition module, an image processing module, a feature identification module and a quality acceptance module. Through machine vision, transparent bottled flaxseed oil is photographed in multiple directions, acquire image features, and complement the color tone. and contrast adjustment, determine whether the impurities are impurities in flaxseed oil, and determine their impact on quality based on acceptance standards.
It realizes accurate detection and identification of impurities in flaxseed oil, ensures the quality and safety of bottled flaxseed oil, and provides consumers with safe choice of cooking oil.
Smart Images
Figure CN119198750A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of edible oil quality detection, and in particular to a system and method for detecting impurities in bottled linseed oil. Background Art
[0002] The oil obtained by pressing flaxseed oil at room temperature is a yellow liquid with a specific smell. Its texture gradually thickens and its color gradually darkens in the air. The quality and safety of agricultural products are directly related to the health and life safety of the people. Agricultural products are an important part of people's daily diet, and their quality and safety directly affect the health of consumers. As a category of agricultural product quality and safety, the safety instructions of flaxseed oil are of great significance to people's safety.
[0003] Impurities in flaxseed oil are mainly caused by incomplete refining process or the separation of natural ingredients under low temperature conditions. Moreover, flaxseed oil has lost most of its nutrients after being stored for a long time, and it tastes bad. It may breed bacteria and produce toxins. Eating it may cause food poisoning, causing symptoms such as nausea, vomiting, and diarrhea. Therefore, it is particularly important to check flaxseed oil before it enters the market. Summary of the invention
[0004] The invention provides a system and method for detecting impurities in bottled linseed oil, which are used to solve the problems raised in the background technology.
[0005] A bottled linseed oil impurity detection system, comprising: An image acquisition module is used to perform multi-directional photography of the transparent bottled linseed oil based on machine vision to obtain multi-directional photographed images; An image processing module, used for performing complementary color toning and contrast enhancement on the multi-directional photographed images based on image attribute values of the multi-directional photographed images to obtain target multi-directional photographed images; A feature identification module, used for obtaining image features of impurity pixels based on multi-directional images of the target, and identifying the image features of the impurity pixels to determine whether the impurities are impurities in linseed oil; The quality acceptance module is used to process the impurities based on the acceptance criteria after determining that the impurities are impurities in the linseed oil, and to determine whether the impurities affect the acceptance quality.
[0006] Preferably, the image acquisition module comprises: A trajectory acquisition unit, used to acquire a moving trajectory of the transparent bottled linseed oil on the quality inspection line, and determine an appearance moving feature trajectory of the transparent bottled linseed oil based on the moving trajectory; A device feature determination unit, used to determine the position features and movement features of each device in the machine vision based on the appearance movement feature trajectory and in combination with the position of the machine vision; The shooting unit is used to shoot the transparent bottled linseed oil in multiple directions using machine vision according to the position characteristics and movement characteristics to obtain multiple direction shooting images.
[0007] Preferably, the image processing module includes: an identification unit, configured to identify the linseed oil part in the bottled linseed oil from the multi-directional photographed image to obtain a first identification result, determine a first accuracy of the first identification result based on a standard feature, and identify the bottle body part from the multi-directional photographed image to obtain a second identification result, and determine a second accuracy of the second identification result based on a standard feature; A first analysis unit is configured to determine a first local feature area based on a difference between the first correctness and a preset correctness, obtain a first local image feature under the first local feature area that affects the correctness from a first recognition result, and determine a first complementary color parameter and a first contrast increase parameter for the multi-directionally shot image based on an image attribute value of the first local image feature; a second analyzing unit, configured to determine a second local feature area based on a difference between the second correctness and a preset correctness, obtain a second local image feature under the second local feature area that affects the correctness from the second recognition result, and determine a second complementary color parameter and a second contrast increase parameter for the multi-directionally shot image based on an image attribute value of the second local image feature; A first correction unit, configured to correct the first complementary color parameters based on the second complementary color parameters to obtain target complementary color parameters; A second correction unit, configured to correct the first contrast increase parameter based on the second contrast increase parameter to obtain a target contrast increase parameter; The processing unit is used to process the multi-directional photographed images according to the target complementary color parameters and contrast increase parameters to obtain the target multi-directional photographed images.
[0008] Preferably, the first correction unit includes: a first parameter determining unit, configured to obtain a first parameter difference between the first complementary color parameter and the second complementary color parameter, and determine a first parameter weight based on the second complementary color parameter; The first parameter correction unit is used to correct the first complementary color parameters based on the product of the first parameter difference and the first parameter weight to obtain target complementary color parameters.
[0009] Preferably, the second correction unit comprises: a second parameter determining unit, configured to obtain a second parameter difference between the first contrast increasing parameter and the second contrast increasing parameter, and determine a second parameter weight based on the second contrast increasing parameter; The second parameter correction unit is used to correct the first contrast increase parameter based on the product of the second parameter difference and the second parameter weight to obtain a target contrast increase parameter.
[0010] Preferably, the feature identification module includes: A panoramic acquisition unit, used to acquire image features corresponding to impurity pixels from the multi-directional images of the target, and mark the image features based on the directional information to obtain an image feature mark group, and extract and acquire impurity image features corresponding to each directional from the image feature mark group, and weight the impurity image features based on the directional information to obtain feature weights, determine a fusion ratio of each impurity image feature based on the feature weights, and fuse all impurity image features according to the fusion ratios to obtain impurity panoramic image features; a feature determination unit, configured to determine a main feature of an impurity based on the impurity panoramic image feature, perform feature quantization on the impurity panoramic image feature to obtain a 512-dimensional binary feature, and determine a reference feature of the impurity from the binary feature; A fusion unit, used for performing feature fusion on the main feature and the reference feature to obtain an image fusion feature, and obtaining an impurity position and an impurity smoothness determined by the image fusion feature; a credibility determination unit, configured to determine a first credibility based on the impurity position, determine a second credibility based on the impurity smoothness, determine a third credibility between the impurity position and the impurity smoothness based on a position-smoothness correlation relationship table of impurities in oil, and determine a comprehensive credibility based on the first credibility, the second credibility and the third credibility; The impurity discrimination unit is used to determine whether the impurity is an impurity in the linseed oil based on the comprehensive credibility.
[0011] Preferably, the impurity discrimination unit comprises: The judgment unit is used to determine that the impurity is an impurity in the linseed oil when the comprehensive credibility is greater than a preset credibility, otherwise, determine that the impurity is an impurity in the bottle.
[0012] Preferably, the quality acceptance module includes: An area judgment unit, used to judge whether the area of the impurity is greater than the area of the acceptance standard; If so, determine whether the impurities affect the acceptance quality; Otherwise, further judgment is made using the impurities; An appearance judgment unit, used to judge whether the appearance characteristics of the impurities meet the appearance characteristics of the acceptance criteria; If so, confirm that the impurities do not affect the acceptance quality; Otherwise, it is determined that the impurity affects the acceptance quality.
[0013] Preferably, the appearance characteristics used by the appearance judgment unit to judge whether the appearance characteristics of the impurities meet the acceptance criteria are specifically: Acquire color characteristics and density distribution characteristics from the appearance characteristics of the impurities, and determine whether the color characteristics and density distribution characteristics meet the color and density requirements of irrelevant impurities in the acceptance criteria; If so, determine that the appearance characteristics of the impurity meet the appearance characteristics of the acceptance criteria; Otherwise, it is determined that the appearance characteristics of the impurity do not meet the appearance characteristics of the acceptance criteria.
[0014] A detection method of a bottled linseed oil impurity detection system, comprising: S1: Based on machine vision, transparent bottled linseed oil is photographed from multiple angles to obtain multi-angle photographed images; S2: Based on the image attribute values of the multi-directional photographed images, complementary coloring and contrast enhancement are performed on the multi-directional photographed images to obtain a target multi-directional photographed image; S3: acquiring image features of impurity pixels based on multi-directional images of the target, and distinguishing the image features of the impurity pixels to determine whether the impurities are impurities in linseed oil; S4: After determining that the impurity is an impurity in the linseed oil, the impurity is processed based on the acceptance criteria to determine whether the impurity affects the acceptance quality.
[0015] Compared with the prior art, the present invention has achieved the following beneficial effects: By taking multi-directional photographs of transparent bottled linseed oil based on machine vision, a multi-directional photographed image is obtained, which provides an image basis for impurity detection of bottled linseed oil. Based on the image attribute values of the multi-directional photographed image, the multi-directional photographed image is toned and the contrast is increased to obtain a target multi-directional photographed image, which provides accurate features for impurity detection of bottled linseed oil. Based on the target multi-directional photographed image, image features of impurity pixels are obtained, and the image features of the impurity pixels are distinguished to determine whether the impurities are impurities in the linseed oil, thereby distinguishing the impurities in the linseed oil. After determining that the impurities are impurities in the linseed oil, the impurities are processed based on the acceptance criteria to determine whether the impurities affect the acceptance quality, thereby ensuring the quality and safety of the bottled linseed oil and providing consumers with a safe choice of edible oil.
[0016] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in this application document.
[0017] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 This is a structural diagram of a system for detecting impurities in bottled linseed oil according to an embodiment of the present invention; Figure 2 is a structural diagram of the image acquisition module described in an embodiment of the present invention; Figure 3 The present invention is a flowchart of a method for detecting impurities in bottled linseed oil according to an embodiment of the present invention. DETAILED DESCRIPTION
[0019] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.
[0020] Embodiment 1: The embodiment of the present invention provides a system for detecting impurities in bottled linseed oil. Figure 1 As shown, including: An image acquisition module is used to perform multi-directional photography of the transparent bottled linseed oil based on machine vision to obtain multi-directional photographed images; An image processing module, used for performing complementary color toning and contrast enhancement on the multi-directional photographed images based on image attribute values of the multi-directional photographed images to obtain target multi-directional photographed images; A feature identification module, used for obtaining image features of impurity pixels based on multi-directional images of the target, and identifying the image features of the impurity pixels to determine whether the impurities are impurities in linseed oil; The quality acceptance module is used to process the impurities based on the acceptance criteria after determining that the impurities are impurities in the linseed oil, and to determine whether the impurities affect the acceptance quality.
[0021] In this embodiment, machine vision includes camera recognition, light source equipment and other auxiliary equipment such as movement and rotation.
[0022] In this embodiment, the image attribute values include resolution, color depth, true color, image size, etc.
[0023] In this embodiment, the acceptance criteria are pre-set according to the production standards of bottled linseed oil.
[0024] The beneficial effects of the above design scheme are as follows: by performing multi-directional shooting of transparent bottled linseed oil based on machine vision, a multi-directional shooting image is obtained, which provides an image basis for impurity detection of bottled linseed oil; based on the image attribute values of the multi-directional shooting image, the multi-directional shooting image is toned and contrast-enhanced to obtain a target multi-directional shooting image, which provides accurate features for impurity detection of bottled linseed oil; based on the target multi-directional shooting image, image features of impurity pixels are obtained, and the image features of the impurity pixels are distinguished to determine whether the impurities are impurities in linseed oil, thereby distinguishing impurities in linseed oil; after determining that the impurities are impurities in linseed oil, the impurities are processed based on the acceptance criteria to determine whether the impurities affect the acceptance quality, thereby ensuring the quality and safety of bottled linseed oil and providing consumers with a safe choice of edible oil.
[0025] Embodiment 2: Based on Example 1, the present invention provides a system for detecting impurities in bottled linseed oil. Figure 2 As shown, the image acquisition module includes: A trajectory acquisition unit, used to acquire a moving trajectory of the transparent bottled linseed oil on the quality inspection line, and determine an appearance moving feature trajectory of the transparent bottled linseed oil based on the moving trajectory; A device feature determination unit, used to determine the position features and movement features of each device in the machine vision based on the appearance movement feature trajectory and in combination with the position of the machine vision; The shooting unit is used to shoot the transparent bottled linseed oil in multiple directions using machine vision according to the position characteristics and movement characteristics to obtain multiple direction shooting images.
[0026] In this embodiment, the position characteristics and movement characteristics of each device in the machine vision are determined in combination with the position of the machine vision, specifically the light source is determined, and the camera moves to the optimal position when the transparent bottle of linseed oil moves, thereby achieving multi-directional shooting and shooting accuracy of the transparent bottle of linseed oil.
[0027] The beneficial effect of the above design scheme is as follows: by obtaining the moving trajectory of the transparent bottled linseed oil on the quality inspection line, and determining the appearance moving feature trajectory of the transparent bottled linseed oil based on the moving trajectory, based on the appearance moving feature trajectory, combined with the position of the machine vision, the position characteristics and movement characteristics of each device in the machine vision are determined, according to the position characteristics and movement characteristics, the transparent bottled linseed oil is photographed in multiple directions using machine vision to obtain a multi-directional photographed image, thereby achieving multi-directional photography and photography accuracy of the transparent bottled linseed oil, and ensuring the comprehensiveness and accuracy of the obtained multi-directional photographed image.
[0028] Embodiment 3: Based on Example 1, the embodiment of the present invention provides a system for detecting impurities in bottled linseed oil, wherein the image processing module comprises: an identification unit, configured to identify the linseed oil part in the bottled linseed oil from the multi-directional photographed image to obtain a first identification result, determine a first accuracy of the first identification result based on a standard feature, and identify the bottle body part from the multi-directional photographed image to obtain a second identification result, and determine a second accuracy of the second identification result based on a standard feature; A first analysis unit is configured to determine a first local feature area based on a difference between the first correctness and a preset correctness, obtain a first local image feature under the first local feature area that affects the correctness from a first recognition result, and determine a first complementary color parameter and a first contrast increase parameter for the multi-directionally shot image based on an image attribute value of the first local image feature; a second analyzing unit, configured to determine a second local feature area based on a difference between the second correctness and a preset correctness, obtain a second local image feature under the second local feature area that affects the correctness from the second recognition result, and determine a second complementary color parameter and a second contrast increase parameter for the multi-directionally shot image based on an image attribute value of the second local image feature; A first correction unit, configured to correct the first complementary color parameters based on the second complementary color parameters to obtain target complementary color parameters; A second correction unit, configured to correct the first contrast increase parameter based on the second contrast increase parameter to obtain a target contrast increase parameter; The processing unit is used to process the multi-directional photographed images according to the target complementary color parameters and contrast increase parameters to obtain the target multi-directional photographed images.
[0029] In this embodiment, the more the first recognition result conforms to the standard feature, the higher the corresponding first accuracy is, and the same is true for the second accuracy.
[0030] In this embodiment, the larger the difference between the first correctness and the preset correctness is, the larger the corresponding first local characteristic area is, and the same is true for the second local characteristic area.
[0031] In this embodiment, the image attribute values include resolution, color depth, true color, image size, etc.
[0032] In this embodiment, the determination of the first complementary color parameter and the first contrast increase parameter is used to compensate for the defects of the image attribute value.
[0033] The beneficial effects of the above design scheme are as follows: a first recognition result is obtained by identifying the linseed oil part in the bottled linseed oil through the multi-directional photographed image, a first accuracy of the first recognition result is determined based on a standard feature, and a second recognition result is obtained by identifying the bottle body part through the multi-directional photographed image, a second accuracy of the second recognition result is determined based on a standard feature, the recognition result is analyzed and processed according to the relationship between the accuracy and the preset accuracy, and the target complementary color parameters and contrast increase parameters are finally determined to process the multi-directional photographed image, thereby obtaining a target multi-directional photographed image, and providing accurate features for impurity detection of bottled linseed oil.
[0034] Embodiment 4: Based on Example 3, the embodiment of the present invention provides a system for detecting impurities in bottled linseed oil, wherein the first correction unit comprises: a first parameter determining unit, configured to obtain a first parameter difference between the first complementary color parameter and the second complementary color parameter, and determine a first parameter weight based on the second complementary color parameter; The first parameter correction unit is used to correct the first complementary color parameters based on the product of the first parameter difference and the first parameter weight to obtain target complementary color parameters.
[0035] The beneficial effect of the above design scheme is: by obtaining the first parameter difference between the first complementary color parameter and the second complementary color parameter, and determining the first parameter weight based on the second complementary color parameter, the first complementary color parameter is corrected based on the product of the first parameter difference and the first parameter weight to obtain the target complementary color parameter, ensuring that the obtained target complementary color parameter can better correct the captured image accurately, and provide accurate features for the detection of impurities in bottled linseed oil.
[0036] Embodiment 5: Based on Example 3, the embodiment of the present invention provides a system for detecting impurities in bottled linseed oil, wherein the second correction unit comprises: a second parameter determining unit, configured to obtain a second parameter difference between the first contrast increasing parameter and the second contrast increasing parameter, and determine a second parameter weight based on the second contrast increasing parameter; The second parameter correction unit is used to correct the first contrast increase parameter based on the product of the second parameter difference and the second parameter weight to obtain a target contrast increase parameter.
[0037] The beneficial effect of the above design scheme is: by obtaining the second parameter difference between the first contrast increase parameter and the second contrast increase parameter, and determining the second parameter weight based on the second contrast increase parameter, the first contrast increase parameter is corrected based on the product of the second parameter difference and the second parameter weight to obtain the target contrast increase parameter, ensuring that the obtained target contrast increase parameter can better accurately correct the captured image, and provide accurate features for bottled linseed oil impurity detection.
[0038] Embodiment 6: Based on Example 1, the embodiment of the present invention provides a bottled linseed oil impurity detection system, wherein the feature identification module comprises: A panoramic acquisition unit, used to acquire image features corresponding to impurity pixels from the multi-directional images of the target, and mark the image features based on the directional information to obtain an image feature mark group, and extract and acquire impurity image features corresponding to each directional from the image feature mark group, and weight the impurity image features based on the directional information to obtain feature weights, determine a fusion ratio of each impurity image feature based on the feature weights, and fuse all impurity image features according to the fusion ratios to obtain impurity panoramic image features; a feature determination unit, configured to determine a main feature of an impurity based on the impurity panoramic image feature, perform feature quantization on the impurity panoramic image feature to obtain a 512-dimensional binary feature, and determine a reference feature of the impurity from the binary feature; A fusion unit, used for performing feature fusion on the main feature and the reference feature to obtain an image fusion feature, and obtaining an impurity position and an impurity smoothness determined by the image fusion feature; a credibility determination unit, configured to determine a first credibility based on the impurity position, determine a second credibility based on the impurity smoothness, determine a third credibility between the impurity position and the impurity smoothness based on a position-smoothness correlation relationship table of impurities in oil, and determine a comprehensive credibility based on the first credibility, the second credibility and the third credibility; The impurity discrimination unit is used to determine whether the impurity is an impurity in the linseed oil based on the comprehensive credibility.
[0039] In this embodiment, the more comprehensive the features of the impurity image in the orientation information are, the greater the corresponding feature weight is.
[0040] In this embodiment, the 512-dimensional binary feature can acquire features from aspects that cannot be observed by the main feature, which is helpful for distinguishing impurities.
[0041] In this embodiment, the position of impurities in the oil-smoothness correlation relationship table is set in advance according to historical acceptance conditions. The greater the correlation, the higher the corresponding third credibility.
[0042] The beneficial effect of the above design scheme is: by fusing different image features to determine the impurity position and impurity smoothness, it is finally determined whether the impurity is an impurity in linseed oil, thereby achieving accurate identification of the impurity and providing a basis for impurity quality inspection.
[0043] Embodiment 7: Based on Example 6, the embodiment of the present invention provides a system for detecting impurities in bottled linseed oil, wherein the impurity identification unit comprises: The judgment unit is used to determine that the impurity is an impurity in the linseed oil when the comprehensive credibility is greater than a preset credibility, otherwise, determine that the impurity is an impurity in the bottle.
[0044] The beneficial effect of the above design is that when the comprehensive credibility is greater than the preset credibility, the impurity is determined to be an impurity in the linseed oil; otherwise, the impurity is determined to be an impurity in the bottle, thereby realizing the judgment of the impurity position.
[0045] Embodiment 8: Based on Example 1, the embodiment of the present invention provides a bottled linseed oil impurity detection system, wherein the quality acceptance module comprises: An area judgment unit, used to judge whether the area of the impurity is greater than the area of the acceptance standard; If so, determine whether the impurities affect the acceptance quality; Otherwise, further judgment is made using the impurities; An appearance judgment unit, used to judge whether the appearance characteristics of the impurities meet the appearance characteristics of the acceptance criteria; If so, confirm that the impurities do not affect the acceptance quality; Otherwise, it is determined that the impurity affects the acceptance quality.
[0046] The beneficial effect of the above design scheme is: after determining that the impurities are impurities in linseed oil, the impurities are processed based on the acceptance criteria to determine whether the impurities affect the acceptance quality, thereby ensuring the quality and safety of bottled linseed oil and providing consumers with a safe edible oil choice.
[0047] Embodiment 9: Based on Example 8, the embodiment of the present invention provides a system for detecting impurities in bottled linseed oil, and the appearance characteristics of the impurities judged by the appearance judgment unit to determine whether the appearance characteristics of the impurities meet the acceptance criteria are specifically: Acquire color characteristics and density distribution characteristics from the appearance characteristics of the impurities, and determine whether the color characteristics and density distribution characteristics meet the color and density requirements of irrelevant impurities in the acceptance criteria; If so, determine that the appearance characteristics of the impurity meet the appearance characteristics of the acceptance criteria; Otherwise, it is determined that the appearance characteristics of the impurity do not meet the appearance characteristics of the acceptance criteria.
[0048] In this example, extraneous impurities are impurities that do not affect the acceptance criteria.
[0049] The beneficial effect of the above design scheme is: by obtaining the color characteristics and density distribution characteristics from the appearance characteristics of the impurities, it is determined whether the color characteristics and density distribution characteristics meet the color and density requirements of irrelevant impurities in the acceptance criteria; if so, it is determined that the appearance characteristics of the impurities meet the appearance characteristics of the acceptance criteria; otherwise, it is determined that the appearance characteristics of the impurities do not meet the appearance characteristics of the acceptance criteria, thereby realizing acceptance judgment of the impurities based on their appearance characteristics.
[0050] Embodiment 10: Based on Example 1, the present invention provides a detection method of a bottled linseed oil impurity detection system, comprising: S1: Based on machine vision, transparent bottled linseed oil is photographed from multiple angles to obtain multi-angle photographed images; S2: Based on the image attribute values of the multi-directional photographed images, complementary coloring and contrast enhancement are performed on the multi-directional photographed images to obtain a target multi-directional photographed image; S3: acquiring image features of impurity pixels based on multi-directional images of the target, and distinguishing the image features of the impurity pixels to determine whether the impurities are impurities in linseed oil; S4: After determining that the impurity is an impurity in the linseed oil, the impurity is processed based on the acceptance criteria to determine whether the impurity affects the acceptance quality.
[0051] In this embodiment, machine vision includes camera recognition, light source equipment and other auxiliary equipment such as movement and rotation.
[0052] In this embodiment, the image attribute values include resolution, color depth, true color, image size, etc.
[0053] In this embodiment, the acceptance criteria are pre-set according to the production standards of bottled linseed oil.
[0054] The beneficial effects of the above design scheme are as follows: by performing multi-directional shooting of transparent bottled linseed oil based on machine vision, a multi-directional shooting image is obtained, which provides an image basis for impurity detection of bottled linseed oil; based on the image attribute values of the multi-directional shooting image, the multi-directional shooting image is toned and contrast-enhanced to obtain a target multi-directional shooting image, which provides accurate features for impurity detection of bottled linseed oil; based on the target multi-directional shooting image, image features of impurity pixels are obtained, and the image features of the impurity pixels are distinguished to determine whether the impurities are impurities in linseed oil, thereby distinguishing impurities in linseed oil; after determining that the impurities are impurities in linseed oil, the impurities are processed based on the acceptance criteria to determine whether the impurities affect the acceptance quality, thereby ensuring the quality and safety of bottled linseed oil and providing consumers with a safe choice of edible oil.
[0055] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of this application document and its equivalent technology, the present invention is also intended to include these changes and variations.
Claims
1. A system for detecting impurities in bottled linseed oil, characterized in that: include: An image acquisition module is used to perform multi-directional photography of the transparent bottled linseed oil based on machine vision to obtain multi-directional photographed images; An image processing module, used for performing complementary color toning and contrast enhancement on the multi-directional photographed images based on image attribute values of the multi-directional photographed images to obtain target multi-directional photographed images; A feature identification module, used for obtaining image features of impurity pixels based on multi-directional images of the target, and identifying the image features of the impurity pixels to determine whether the impurities are impurities in linseed oil; The quality acceptance module is used to process the impurities based on the acceptance criteria after determining that the impurities are impurities in the linseed oil, and to determine whether the impurities affect the acceptance quality.
2. A bottled linseed oil impurity detection system according to claim 1, characterized in that: The image acquisition module comprises: A trajectory acquisition unit, used to acquire a moving trajectory of the transparent bottled linseed oil on the quality inspection line, and determine an appearance moving feature trajectory of the transparent bottled linseed oil based on the moving trajectory; A device feature determination unit, used to determine the position features and movement features of each device in the machine vision based on the appearance movement feature trajectory and in combination with the position of the machine vision; The shooting unit is used to shoot the transparent bottled linseed oil in multiple directions using machine vision according to the position characteristics and movement characteristics to obtain multiple direction shooting images.
3. The impurity detection system for bottled linseed oil according to claim 1, characterized in that: The image processing module comprises: an identification unit, configured to identify the linseed oil part in the bottled linseed oil from the multi-directional photographed image to obtain a first identification result, determine a first accuracy of the first identification result based on a standard feature, and identify the bottle body part from the multi-directional photographed image to obtain a second identification result, and determine a second accuracy of the second identification result based on a standard feature; A first analysis unit is configured to determine a first local feature area based on a difference between the first correctness and a preset correctness, obtain a first local image feature under the first local feature area that affects the correctness from a first recognition result, and determine a first complementary color parameter and a first contrast increase parameter for the multi-directionally shot image based on an image attribute value of the first local image feature; a second analyzing unit, configured to determine a second local feature area based on a difference between the second correctness and a preset correctness, obtain a second local image feature under the second local feature area that affects the correctness from the second recognition result, and determine a second complementary color parameter and a second contrast increase parameter for the multi-directionally shot image based on an image attribute value of the second local image feature; A first correction unit, configured to correct the first complementary color parameters based on the second complementary color parameters to obtain target complementary color parameters; A second correction unit, configured to correct the first contrast increase parameter based on the second contrast increase parameter to obtain a target contrast increase parameter; The processing unit is used to process the multi-directional photographed images according to the target complementary color parameters and contrast increase parameters to obtain the target multi-directional photographed images.
4. A bottled linseed oil impurity detection system according to claim 3, characterized in that: The first correction unit comprises: a first parameter determining unit, configured to obtain a first parameter difference between the first complementary color parameter and the second complementary color parameter, and determine a first parameter weight based on the second complementary color parameter; The first parameter correction unit is used to correct the first complementary color parameters based on the product of the first parameter difference and the first parameter weight to obtain target complementary color parameters.
5. The impurity detection system for bottled linseed oil according to claim 3, characterized in that: The second correction unit comprises: a second parameter determining unit, configured to obtain a second parameter difference between the first contrast increasing parameter and the second contrast increasing parameter, and determine a second parameter weight based on the second contrast increasing parameter; The second parameter correction unit is used to correct the first contrast increase parameter based on the product of the second parameter difference and the second parameter weight to obtain a target contrast increase parameter.
6. The impurity detection system for bottled linseed oil according to claim 1, characterized in that: The feature identification module comprises: A panoramic acquisition unit, used to acquire image features corresponding to impurity pixels from the multi-directional images of the target, and mark the image features based on the directional information to obtain an image feature mark group, and extract and acquire impurity image features corresponding to each directional from the image feature mark group, and weight the impurity image features based on the directional information to obtain feature weights, determine a fusion ratio of each impurity image feature based on the feature weights, and fuse all impurity image features according to the fusion ratios to obtain impurity panoramic image features; a feature determination unit, configured to determine a main feature of an impurity based on the impurity panoramic image feature, perform feature quantization on the impurity panoramic image feature to obtain a 512-dimensional binary feature, and determine a reference feature of the impurity from the binary feature; A fusion unit, used for performing feature fusion on the main feature and the reference feature to obtain an image fusion feature, and obtaining an impurity position and an impurity smoothness determined by the image fusion feature; a credibility determination unit, configured to determine a first credibility based on the impurity position, determine a second credibility based on the impurity smoothness, determine a third credibility between the impurity position and the impurity smoothness based on a position-smoothness correlation relationship table of impurities in oil, and determine a comprehensive credibility based on the first credibility, the second credibility and the third credibility; The impurity discrimination unit is used to determine whether the impurity is an impurity in the linseed oil based on the comprehensive credibility.
7. A bottled linseed oil impurity detection system according to claim 6, characterized in that: The impurity discrimination unit comprises: The judgment unit is used to determine that the impurity is an impurity in the linseed oil when the comprehensive credibility is greater than a preset credibility, otherwise, determine that the impurity is an impurity in the bottle.
8. The system for detecting impurities in bottled linseed oil according to claim 1, characterized in that: The quality acceptance module includes: An area judgment unit, used to judge whether the area of the impurity is greater than the area of the acceptance standard; If so, determine whether the impurities affect the acceptance quality; Otherwise, further judgment is made using the impurities; An appearance judgment unit, used to judge whether the appearance characteristics of the impurities meet the appearance characteristics of the acceptance criteria; If so, confirm that the impurities do not affect the acceptance quality; Otherwise, it is determined that the impurity affects the acceptance quality.
9. The system for detecting impurities in bottled linseed oil according to claim 8, characterized in that: The appearance characteristics used by the appearance judgment unit to judge whether the appearance characteristics of the impurities meet the acceptance criteria are specifically: Acquire color characteristics and density distribution characteristics from the appearance characteristics of the impurities, and determine whether the color characteristics and density distribution characteristics meet the color and density requirements of irrelevant impurities in the acceptance criteria; If so, determine that the appearance characteristics of the impurity meet the appearance characteristics of the acceptance criteria; Otherwise, it is determined that the appearance characteristics of the impurity do not meet the appearance characteristics of the acceptance criteria.
10. The detection method of the bottled linseed oil impurity detection system according to claim 1, characterized in that: include: S1: Based on machine vision, transparent bottled linseed oil is photographed from multiple angles to obtain multi-angle photographed images; S2: Based on the image attribute values of the multi-directional photographed images, complementary coloring and contrast enhancement are performed on the multi-directional photographed images to obtain a target multi-directional photographed image; S3: acquiring image features of impurity pixels based on multi-directional images of the target, and distinguishing the image features of the impurity pixels to determine whether the impurities are impurities in linseed oil; S4: After determining that the impurity is an impurity in the linseed oil, the impurity is processed based on the acceptance criteria to determine whether the impurity affects the acceptance quality.