Intelligent detection method and device for wine with different covers and computer equipment

Abstract

The invention relates to an intelligent detection method and device for different-cover wine and computer equipment. The method comprises the following steps: obtaining a bottle cap image of each wine bottle, and carrying out image preprocessing on the bottle cap image of each wine bottle to obtain a target bottle cap image of each wine bottle; extracting feature information of each target bottle cap image through a feature extraction network based on the target bottle cap image of each wine bottle, and identifying abnormal information of each target bottle cap image through a feature comparison strategy based on the feature information of each target bottle cap image; and on the basis of the abnormal information of each target bottle cap image, through a different-cap wine evaluation strategy, identifying the different-cap wine probability of the different-cap wine type corresponding to each target bottle cap image, and on the basis of the different-cap wine probability of the different-cap wine type corresponding to each target bottle cap image, identifying the information of each abnormal wine bottle. By adopting the method, the detection efficiency of different-cap wine can be improved.

Description

Technical Field

[0001] The present application relates to the field of image detection and image processing technology, and in particular to an intelligent detection method, device and computer equipment for wine with abnormal caps. Background Art

[0002] To prevent wine with foreign caps from entering the market, bottle cap inspection is essential on wine packaging lines. Traditional inspection methods rely on manual visual inspection, but suffer from low accuracy and efficiency. This is especially true when bottle caps are similar in color, making it easy to miss detection. This can lead to wine with foreign caps entering the market, triggering consumer complaints and damaging brand reputation. Therefore, improving the accuracy and efficiency of wine with foreign caps detection is a current research priority.

[0003] In traditional technology, imported FBI equipment is installed behind the wine machine, but it is expensive and costly. Moreover, the efficiency of detecting wine with foreign caps using this method is low and cannot meet the needs of high-frequency and efficient wine production line processes, resulting in poor detection efficiency of wine with foreign caps. Summary of the Invention

[0004] Based on this, it is necessary to provide an intelligent detection method, device, computer equipment, computer-readable storage medium and computer program product for wine with abnormal caps in order to solve the above technical problems.

[0005] In a first aspect, the present application provides an intelligent detection method for wine with an abnormal cap, comprising:

[0006] Acquire a bottle cap image of each wine bottle, and perform image preprocessing on the bottle cap image of each wine bottle to obtain a target bottle cap image of each wine bottle;

[0007] Based on the target bottle cap images of each of the wine bottles, feature information of each of the target bottle cap images is extracted through a feature extraction network, and based on the feature information of each of the target bottle cap images, abnormal information of each of the target bottle cap images is identified through a feature comparison strategy;

[0008] Based on the abnormal information of each target bottle cap image, the abnormal wine evaluation strategy is used to identify the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image, and based on the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image, the information of each abnormal wine bottle is identified.

[0009] Optionally, performing image preprocessing on the bottle cap images of the wine bottles to obtain target bottle cap images of the wine bottles includes:

[0010] For each wine bottle, split the bottle cap image into sub-bottle cap images corresponding to each photographing condition;

[0011] By using the image optimization strategy corresponding to each of the photographing conditions, the sub-bottle cap images corresponding to each of the photographing conditions are optimized to obtain the target sub-bottle cap images corresponding to each of the photographing conditions;

[0012] The target sub-bottle cap images corresponding to all photographing conditions are used as the target bottle cap image of the wine bottle.

[0013] Optionally, the feature extraction network includes an image feature extraction network corresponding to each of the photographing conditions and a character feature extraction network corresponding to each of the photographing conditions; extracting feature information of each of the target bottle cap images based on the target bottle cap images of each of the wine bottles through the feature extraction network includes:

[0014] For each wine bottle, based on the target sub-bottle cap image corresponding to each of the photographing conditions, the image feature data corresponding to each of the photographing conditions is extracted through the image feature extraction network of each of the photographing conditions;

[0015] Based on the target sub-bottle cap image corresponding to each of the photographing conditions, extracting the character feature data corresponding to each of the photographing conditions through the character feature extraction network of each of the photographing conditions;

[0016] The image feature data corresponding to each of the photographing conditions and the character feature data corresponding to each of the photographing conditions are used as feature information of the target bottle cap image.

[0017] Optionally, identifying abnormal information of each target bottle cap image through a feature comparison strategy based on feature information of each target bottle cap image includes:

[0018] Acquiring sample feature information of a sample wine bottle and sample image data of the sample wine bottle;

[0019] For each wine bottle, based on the feature information of the target bottle cap image of the wine bottle and the sample feature information of the sample wine bottle, a feature similarity recognition network is used to identify the feature difference information of the wine bottle;

[0020] identifying image difference information of the wine bottle based on the sample image data of the sample wine bottle and the target image data of the wine bottle;

[0021] The feature difference information of the wine bottle and the image difference information of the wine bottle are used as abnormal information of the target image data corresponding to the wine bottle.

[0022] Optionally, the step of identifying the probability of an alien-cap wine of an alien-cap wine type corresponding to each target bottle cap image based on the abnormal information of each target bottle cap image by using an alien-cap wine evaluation strategy includes:

[0023] For each wine bottle, identifying the type of wine with a different cap based on the image difference information of the wine bottle;

[0024] Based on the characteristic difference information of the wine bottle, the probability of the wine with different caps in the target bottle cap image corresponding to the wine bottle is identified by using the probability identification strategy of the wine with different caps of the wine with different caps type.

[0025] Optionally, the identifying of each abnormal wine bottle information based on the probability of the wine having an abnormal cap corresponding to each target bottle cap image includes:

[0026] Obtaining a probability threshold value of wine with different caps for each type of wine with different caps;

[0027] Based on the probability threshold of each wine type with different caps and the probability of each wine type with different caps corresponding to each target bottle cap image, screening abnormal bottle cap images of each wine type with different caps in each target bottle cap image;

[0028] The type of wine with an abnormal cap corresponding to each abnormal bottle cap image and the probability of wine with an abnormal cap corresponding to each abnormal bottle cap image are used as information of each abnormal wine bottle.

[0029] In a second aspect, the present application also provides an intelligent detection device for wine with an abnormal cover, comprising:

[0030] an acquisition module, configured to acquire the bottle cap image of each wine bottle and perform image preprocessing on the bottle cap image of each wine bottle to obtain a target bottle cap image of each wine bottle;

[0031] an extraction module for extracting feature information of each target bottle cap image based on the target bottle cap image of each wine bottle through a feature extraction network, and identifying abnormal information of each target bottle cap image through a feature comparison strategy based on the feature information of each target bottle cap image;

[0032] The recognition module is used to identify the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image based on the abnormal information of each target bottle cap image through the abnormal wine evaluation strategy, and identify the information of each abnormal wine bottle based on the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image.

[0033] Optionally, the acquisition module is specifically configured to:

[0034] For each wine bottle, split the bottle cap image into sub-bottle cap images corresponding to each photographing condition;

[0035] By using the image optimization strategy corresponding to each of the photographing conditions, the sub-bottle cap images corresponding to each of the photographing conditions are optimized to obtain the target sub-bottle cap images corresponding to each of the photographing conditions;

[0036] The target sub-bottle cap images corresponding to all photographing conditions are used as the target bottle cap image of the wine bottle.

[0037] Optionally, the extraction module is specifically configured to:

[0038] For each wine bottle, based on the target sub-bottle cap image corresponding to each of the photographing conditions, the image feature data corresponding to each of the photographing conditions is extracted through the image feature extraction network of each of the photographing conditions;

[0039] Based on the target sub-bottle cap image corresponding to each of the photographing conditions, extracting the character feature data corresponding to each of the photographing conditions through the character feature extraction network of each of the photographing conditions;

[0040] The image feature data corresponding to each of the photographing conditions and the character feature data corresponding to each of the photographing conditions are used as feature information of the target bottle cap image.

[0041] Optionally, the extraction module is specifically configured to:

[0042] Acquiring sample feature information of a sample wine bottle and sample image data of the sample wine bottle;

[0043] For each wine bottle, based on the feature information of the target bottle cap image of the wine bottle and the sample feature information of the sample wine bottle, a feature similarity recognition network is used to identify the feature difference information of the wine bottle;

[0044] identifying image difference information of the wine bottle based on the sample image data of the sample wine bottle and the target image data of the wine bottle;

[0045] The feature difference information of the wine bottle and the image difference information of the wine bottle are used as abnormal information of the target image data corresponding to the wine bottle.

[0046] Optionally, the identification module is specifically configured to:

[0047] For each wine bottle, identifying the type of wine with a different cap based on the image difference information of the wine bottle;

[0048] Based on the characteristic difference information of the wine bottle, the probability of the wine with different caps in the target bottle cap image corresponding to the wine bottle is identified by using the probability identification strategy of the wine with different caps of the wine with different caps type.

[0049] Optionally, the identification module is specifically configured to:

[0050] Obtaining a probability threshold value of wine with different caps for each type of wine with different caps;

[0051] Based on the probability threshold of each wine type with different caps and the probability of each wine type with different caps corresponding to each target bottle cap image, screening abnormal bottle cap images of each wine type with different caps in each target bottle cap image;

[0052] The type of wine with an abnormal cap corresponding to each abnormal bottle cap image and the probability of wine with an abnormal cap corresponding to each abnormal bottle cap image are used as information of each abnormal wine bottle.

[0053] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the methods described in the first aspect when executing the computer program.

[0054] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of any one of the methods in the first aspect.

[0055] In a fifth aspect, the present application provides a computer program product, wherein the computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of any one of the methods in the first aspect are implemented.

[0056] The above-mentioned intelligent detection method, device and computer equipment for abnormal wine caps obtain the bottle cap image of each wine bottle and perform image preprocessing on the bottle cap image of each wine bottle to obtain the target bottle cap image of each wine bottle; based on the target bottle cap image of each wine bottle, the feature information of each target bottle cap image is extracted through a feature extraction network, and based on the feature information of each target bottle cap image, the abnormal information of each target bottle cap image is identified through a feature comparison strategy; based on the abnormal information of each target bottle cap image, the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image is identified through an abnormal wine evaluation strategy, and based on the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image, the information of each abnormal wine bottle is identified. This solution combines the acquired bottle cap images of each wine bottle to achieve a unified acquisition and separate processing technical solution, effectively improving the detection efficiency of each wine bottle. Then, this solution processes the acquired images to avoid environmental interference, noise interference, and shooting problems with weak image feature information. Therefore, no complex equipment is required for image acquisition and clear feature representation of image information can be obtained. Then, after extracting features from each target bottle cap image, this solution conducts a comprehensive and comprehensive detection of the bottle cap information of the wine bottle through strategies such as feature comparison and evaluation of wine with different caps, thereby ensuring the accuracy of identifying wine with different caps. This allows this solution to simultaneously detect multiple wine bottles while ensuring the accuracy of identifying wine with different caps, thereby comprehensively improving the detection efficiency of wine with different caps. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Figure 1 Schematic diagram of a flow chart of an intelligent detection method for wine with a different cover in one embodiment;

[0059] Figure 2 Schematic diagram of a process for intelligent detection of wine with different caps in one embodiment;

[0060] Figure 3 1. It is a structural block diagram of an intelligent detection device for wine with abnormal cover in one embodiment;

[0061] Figure 4 FIG. 1 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0062] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0063] The intelligent detection method for wine with foreign caps provided in the embodiments of the present application can be applied to an intelligent detection system for wine with foreign caps. The system can be applied to a terminal, which can be, but is not limited to, various personal computers, laptops, mid-range computers, etc. The terminal combines the acquired bottle cap images of each wine bottle to achieve a unified acquisition and separate processing technical solution, effectively improving the detection efficiency of each wine bottle. This solution then processes the acquired images to avoid environmental interference, noise interference, and weak image feature information representation. This eliminates the need for complex equipment for image acquisition and allows for the acquisition of clear feature representation image information. After feature extraction from each target bottle cap image, this solution conducts a comprehensive and comprehensive detection of the bottle cap information through feature comparison, foreign cap evaluation, and other strategies, thereby ensuring the accuracy of wine with foreign caps. This allows the solution to simultaneously detect multiple wine bottles while ensuring the accuracy of wine with foreign caps, thereby comprehensively improving the detection efficiency of wine with foreign caps.

[0064] In an exemplary embodiment, Figure 1 As shown, a smart detection method for wine with a different cover is provided, which is described by taking the application of the method to a terminal as an example, and includes the following steps S101 to S103.

[0065] Step S101 : acquiring a bottle cap image of each wine bottle, and performing image preprocessing on the bottle cap image of each wine bottle to obtain a target bottle cap image of each wine bottle.

[0066] In this embodiment, a terminal uses an intelligent camera to capture images of different wine bottle caps using different colors, light sources, and angles, obtaining images of each bottle cap. The intelligent camera uses a 16-megapixel industrial camera with an adjustable zoom lens with a focal length of 16mm-50mm. The light source is an LED with a brightness range of 1000-5000 lux, which automatically adjusts according to ambient lighting conditions. The image resolution is 16 megapixels. Shooting angles include the top view and the side view (surround view) of the bottle cap. The terminal then performs image preprocessing on each bottle cap image to obtain a target cap image. This preprocessing includes image denoising using a median filter algorithm and image enhancement using a histogram equalization algorithm. Image optimization is also included; the specific optimization methods will be described in detail later.

[0067] Step S102 , based on the target bottle cap image of each wine bottle, feature information of each target bottle cap image is extracted through a feature extraction network, and based on the feature information of each target bottle cap image, abnormal information of each target bottle cap image is identified through a feature comparison strategy.

[0068] In this embodiment, the terminal uses a feature extraction network to extract feature information from target bottle cap images of each wine bottle. Based on this feature information, the terminal then uses a feature comparison strategy to identify abnormalities in each target bottle cap image. The feature extraction network includes an image feature extraction network corresponding to each photographic condition and a character feature extraction network corresponding to each photographic condition. The image feature extraction network is a convolutional neural network based on an attention mechanism and an RGB color space model for color feature extraction. It extracts the average value of the R, G, and B channels of the bottle cap image as the color feature. The character feature extraction network is a neural network that utilizes an optical character recognition algorithm. Each photographic condition corresponds to a different photographic brightness, color, or angle. Examples include red photographic conditions, blue photographic conditions, yellow photographic conditions, 1000-2000 lux photographic conditions, 2000-3000 lux photographic conditions, photographic conditions for the bottle cap surface, and photographic conditions for the bottle cap side. The abnormality information includes feature difference information for the wine bottle and image difference information for the wine bottle. The specific extraction process will be described in detail later.

[0069] Step S103, based on the abnormal information of each target bottle cap image, the probability of the abnormal wine of the abnormal cap type corresponding to each target bottle cap image is identified through the abnormal wine evaluation strategy, and based on the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image, the information of each abnormal wine bottle is identified.

[0070] In this embodiment, based on the abnormal information of each target bottle cap image and the abnormal wine evaluation strategy, the terminal identifies the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image. Furthermore, based on the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image, the terminal identifies the information of each abnormal wine bottle. The abnormal wine type includes, but is not limited to, image abnormality type, color abnormality type, anti-counterfeiting abnormality type, shape abnormality type, or material abnormality type. The abnormal wine bottle information includes the abnormal wine type corresponding to each abnormal bottle cap image and the abnormal wine probability corresponding to each abnormal bottle cap image. The specific identification process will be described in detail later.

[0071] Based on the above scheme, by combining the acquired bottle cap images of each wine bottle, a technical scheme of unified acquisition and separate processing can be achieved, which effectively improves the detection efficiency of each wine bottle. Then, this scheme avoids environmental interference, noise interference, and shooting problems with weak image feature information representation by performing image processing on the acquired images, thereby obtaining clear feature representation image information without the need for complex equipment for image acquisition. Then, after extracting features from each target bottle cap image, this scheme conducts a comprehensive and comprehensive detection of the bottle cap information of the wine bottle through strategies such as feature comparison and evaluation of wine with different caps, thereby ensuring the accuracy of identifying wine with different caps. This allows this scheme to simultaneously detect multiple wine bottles while ensuring the accuracy of identifying wine with different caps, thereby comprehensively improving the detection efficiency of wine with different caps.

[0072] Optionally, the bottle cap images of each wine bottle are subjected to image preprocessing to obtain target bottle cap images of each wine bottle, including: for each wine bottle, splitting the bottle cap image of the wine bottle into sub-bottle cap images corresponding to each photographing condition; performing image optimization processing on the sub-bottle cap images corresponding to each photographing condition through an image optimization strategy corresponding to each photographing condition to obtain target sub-bottle cap images corresponding to each photographing condition; and using the target sub-bottle cap images corresponding to all photographing conditions as the target bottle cap image of the wine bottle.

[0073] In this embodiment, for each wine bottle, the terminal splits the bottle cap image into sub-bottle cap images corresponding to each photographic condition. Then, image processing is performed on each sub-bottle cap image using image denoising using a median filter algorithm and image enhancement using a histogram equalization algorithm. Image optimization strategies corresponding to each photographic condition are then used to optimize the sub-bottle cap images corresponding to each photographic condition, thereby obtaining target sub-bottle cap images corresponding to each photographic condition. Specifically, the terminal trains a deep convolution-based convolutional neural network using sample training images for different photographic conditions and target training images corresponding to each sample training image, thereby obtaining an image optimization network for each photographic condition. The terminal then uses this image optimization network to optimize the sub-bottle cap images corresponding to each photographic condition, thereby obtaining target sub-bottle cap images corresponding to each photographic condition.

[0074] Finally, the terminal uses the target sub-bottle cap images corresponding to all the shooting conditions as the target bottle cap image of the wine bottle.

[0075] Based on the above scheme, by combining interference removal and image enhancement processing, the image is optimized according to different shooting conditions, thereby eliminating the image shooting anomalies and feature fuzziness of traditional camera equipment, effectively improving the feature representation clarity of each target bottle cap image, and the image effect of each target bottle cap image, thereby effectively improving the image representation effect of the target bottle cap image.

[0076] Optionally, the feature extraction network includes an image feature extraction network corresponding to each photographing condition, and a character feature extraction network corresponding to each photographing condition; based on the target bottle cap image of each wine bottle, feature information of each target bottle cap image is extracted through the feature extraction network, including: for each wine bottle, based on the target sub-bottle cap image corresponding to each photographing condition, image feature data corresponding to each photographing condition is extracted through the image feature extraction network of each photographing condition; based on the target sub-bottle cap image corresponding to each photographing condition, character feature data corresponding to each photographing condition is extracted through the character feature extraction network of each photographing condition; the image feature data corresponding to each photographing condition and the character feature data corresponding to each photographing condition are used as feature information of the target bottle cap image.

[0077] In this embodiment, for each wine bottle, the terminal uses an image feature extraction network based on the target sub-bottle cap image corresponding to each photographic condition to extract image feature data corresponding to each photographic condition. Based on the target sub-bottle cap image corresponding to each photographic condition, the terminal uses a character feature extraction network based on the target sub-bottle cap image to extract character feature data corresponding to each photographic condition. This image feature extraction network includes both an RGB color space model for color feature extraction and a convolutional neural network (using an attention mechanism) for feature extraction of image shape, structure, layout, and other image features.

[0078] Finally, the terminal uses the image feature data corresponding to each photographing condition and the character feature data corresponding to each photographing condition as feature information of the target bottle cap image.

[0079] Based on the above solution, image features are extracted from three perspectives: image features, color features, and character features, which improves the comprehensiveness of feature extraction of the target bottle cap image.

[0080] Optionally, based on the feature information of each target bottle cap image, abnormal information of each target bottle cap image is identified through a feature comparison strategy, including: obtaining sample feature information of a sample wine bottle and sample image data of the sample wine bottle; for each wine bottle, based on the feature information of the target bottle cap image of the wine bottle and the sample feature information of the sample wine bottle, identifying feature difference information of the wine bottle through a feature similarity recognition network; based on the sample image data of the sample wine bottle and the target image data of the wine bottle, identifying image difference information of the wine bottle; and using the feature difference information of the wine bottle and the image difference information of the wine bottle as abnormal information of the target image data corresponding to the wine bottle.

[0081] In this embodiment, the terminal obtains sample feature information of the sample wine bottle and sample image data of the sample wine bottle, wherein the sample feature information includes sample image feature data corresponding to each photographing condition and sample character feature data corresponding to each photographing condition.

[0082] For each bottle, the terminal calculates the image feature similarity and character feature similarity between the target bottle cap image and the sample bottle under various photo conditions using a feature similarity recognition network based on the feature information of the target bottle cap image and the sample bottle. The terminal then uses the feature information corresponding to the image feature similarity or character feature similarity below a preset similarity threshold as the bottle's feature difference information. The feature similarity recognition network is a neural network based on the cosine similarity algorithm.

[0083] Based on the sample image data of the sample wine bottle and the target image data of the wine bottle, the image difference information of the wine bottle is identified through an image difference recognition network. The image difference recognition network is based on a deep learning convolutional neural network.

[0084] Finally, the terminal uses the feature difference information of the wine bottle and the image difference information of the wine bottle as abnormal information of the target image data corresponding to the wine bottle.

[0085] Based on the above solution, the abnormal information of wine bottles is identified from two perspectives: feature difference and image difference, which improves the comprehensiveness of the abnormal information identification of wine bottles.

[0086] Optionally, based on the abnormal information of each target bottle cap image, the probability of the wine with an unusual cap corresponding to the unusual cap type of each target bottle cap image is identified through an unusual cap wine evaluation strategy, including: for each wine bottle, based on the image difference information of the wine bottle, identifying the unusual cap wine type of the wine bottle; based on the feature difference information of the wine bottle, through an unusual cap wine probability identification strategy for the unusual cap wine type, identifying the probability of the wine with an unusual cap corresponding to the target bottle cap image of the wine bottle.

[0087] In this embodiment, the terminal identifies the type of wine with an altered cap for each wine bottle based on the bottle's image difference information. Based on the bottle's characteristic difference information, the terminal then uses a probability identification strategy for altered cap wine types to identify the probability of an altered cap wine in the target bottle cap image corresponding to the bottle. The terminal presets a correspondence between different types of wine with altered caps and image difference ranges. Based on this correspondence, the terminal then identifies the image difference range that corresponds to the bottle's image difference information, thereby determining the type of wine with an altered cap corresponding to the bottle. The implementation of the probability identification strategy for altered cap wine involves the terminal presetting characteristic difference ranges corresponding to different probability values ​​for each type of wine with an altered cap. The terminal then identifies the probability value for an altered cap wine corresponding to the bottle based on the characteristic difference range to which the bottle's characteristic information falls.

[0088] Based on the above scheme, by combining image difference information and feature difference information, the type and range of wine with different caps corresponding to the wine bottle are comprehensively judged, thereby improving the comprehensiveness and accuracy of the recognition of the different cap information of the wine bottle.

[0089] Optionally, based on the probability of different-cap wine of the different-cap wine type corresponding to each target bottle cap image, information of each abnormal wine bottle is identified, including: obtaining a probability threshold of different-cap wine of each different-cap wine type; screening abnormal bottle cap images of each different-cap wine type in each target bottle cap image based on the probability threshold of different-cap wine of each different-cap wine type and the probability of different-cap wine of the different-cap wine type corresponding to each target bottle cap image; and using the different-cap wine type corresponding to each abnormal bottle cap image and the probability of different-cap wine corresponding to each abnormal bottle cap image as information of each abnormal wine bottle.

[0090] In this embodiment, the terminal obtains a probability threshold for each type of wine with an abnormal cap. Then, based on the probability threshold for each type of wine with an abnormal cap and the probability of the abnormal cap corresponding to each target bottle cap image, the terminal filters each target bottle cap image for abnormal bottle cap images of each type of wine with an abnormal cap. Finally, the terminal uses the abnormal bottle cap type and the probability of the abnormal cap corresponding to each abnormal bottle cap image as information about each abnormal bottle.

[0091] Based on the above scheme, this scheme can not only identify wine with strange caps, but also identify the corresponding type of wine with strange caps, thereby improving the comprehensiveness of the recognition of wine with strange caps. In addition, in the actual experimental process, different algorithms of this scheme run independently, and the overall algorithm processing time is controlled at about 50 milliseconds. At a speed of 12.5 bottles per second, the accuracy rate can reach more than 99.9%. This not only improves the detection efficiency of wine with strange caps to fully meet the needs of high-speed production lines, but also effectively prevents wine with strange caps from entering the market, reduces consumer complaints, and enhances brand reputation.

[0092] The application also provides an example of intelligent detection of wine with different caps, such as Figure 2 As shown, the specific processing process includes the following steps:

[0093] Step S201: Acquire the bottle cap image of each wine bottle.

[0094] Step S202 : For each wine bottle, the bottle cap image is split into sub-bottle cap images corresponding to each photographing condition.

[0095] Step S203 , performing image optimization processing on the sub-bottle cap images corresponding to the respective photographing conditions by using the image optimization strategies corresponding to the respective photographing conditions, and obtaining target sub-bottle cap images corresponding to the respective photographing conditions.

[0096] Step S204 , taking the target sub-bottle cap images corresponding to all the photographing conditions as the target bottle cap image of the wine bottle.

[0097] Step S205 : For each wine bottle, based on the target sub-bottle cap image corresponding to each photographing condition, the image feature data corresponding to each photographing condition is extracted through the image feature extraction network of each photographing condition.

[0098] Step S206 , based on the target sub-bottle cap image corresponding to each photographing condition, character feature data corresponding to each photographing condition is extracted through a character feature extraction network of each photographing condition.

[0099] In step S207 , the image feature data corresponding to each photographing condition and the character feature data corresponding to each photographing condition are used as feature information of the target bottle cap image.

[0100] Step S208: Acquire sample feature information of the sample wine bottle and sample image data of the sample wine bottle.

[0101] Step S209 : For each wine bottle, based on the feature information of the target bottle cap image and the sample feature information of the sample wine bottle, a feature similarity recognition network is used to identify the feature difference information of the wine bottle.

[0102] Step S210 , identifying image difference information of the wine bottle based on the sample image data of the sample wine bottle and the target image data of the wine bottle.

[0103] In step S211 , the feature difference information of the wine bottle and the image difference information of the wine bottle are used as abnormal information of the target image data corresponding to the wine bottle.

[0104] Step S212: for each wine bottle, identify the type of wine with different caps based on the image difference information of the wine bottle.

[0105] Step S213 , based on the feature difference information of the wine bottle, the probability of the target bottle cap image corresponding to the wine bottle is identified by using the probability identification strategy of the wine with different caps of the wine with different caps type.

[0106] Step S214, obtaining the probability threshold of different-cover wine for each different-cover wine type.

[0107] Step S215 , based on the probability threshold of wine with different caps for each wine with different caps and the probability of wine with different caps for each wine with different caps corresponding to each target bottle cap image, filter out abnormal bottle cap images of each wine with different caps in each target bottle cap image.

[0108] In step S216, the type of wine with an abnormal cap corresponding to each abnormal bottle cap image and the probability of wine with an abnormal cap corresponding to each abnormal bottle cap image are used as information of each abnormal wine bottle.

[0109] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0110] Based on the same inventive concept, embodiments of the present application also provide an intelligent detection device for wine with different caps, which is used to implement the intelligent detection method for wine with different caps described above. The solution provided by this device is similar to the solution described in the method described above. Therefore, the specific limitations of one or more embodiments of the intelligent detection device for wine with different caps provided below can be found in the above-mentioned limitations of the intelligent detection method for wine with different caps, and will not be repeated here.

[0111] In an exemplary embodiment, Figure 3 As shown, an intelligent detection device for wine with a different cover is provided, comprising: an acquisition module 310, an extraction module 320 and an identification module 330, wherein:

[0112] An acquisition module 310 is configured to acquire a bottle cap image of each wine bottle and perform image preprocessing on the bottle cap image of each wine bottle to obtain a target bottle cap image of each wine bottle;

[0113] An extraction module 320 is configured to extract feature information of each target bottle cap image of each wine bottle through a feature extraction network, and identify abnormal information of each target bottle cap image through a feature comparison strategy based on the feature information of each target bottle cap image;

[0114] The identification module 330 is used to identify the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image based on the abnormal information of each target bottle cap image through the abnormal wine evaluation strategy, and identify the information of each abnormal wine bottle based on the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image.

[0115] Optionally, the acquisition module 310 is specifically configured to:

[0116] For each wine bottle, split the bottle cap image into sub-bottle cap images corresponding to each photographing condition;

[0117] By using the image optimization strategy corresponding to each of the photographing conditions, the sub-bottle cap images corresponding to each of the photographing conditions are optimized to obtain the target sub-bottle cap images corresponding to each of the photographing conditions;

[0118] The target sub-bottle cap images corresponding to all photographing conditions are used as the target bottle cap image of the wine bottle.

[0119] Optionally, the extraction module 320 is specifically configured to:

[0120] For each wine bottle, based on the target sub-bottle cap image corresponding to each of the photographing conditions, the image feature data corresponding to each of the photographing conditions is extracted through the image feature extraction network of each of the photographing conditions;

[0121] Based on the target sub-bottle cap image corresponding to each of the photographing conditions, extracting the character feature data corresponding to each of the photographing conditions through the character feature extraction network of each of the photographing conditions;

[0122] The image feature data corresponding to each of the photographing conditions and the character feature data corresponding to each of the photographing conditions are used as feature information of the target bottle cap image.

[0123] Optionally, the extraction module 320 is specifically configured to:

[0124] Acquiring sample feature information of a sample wine bottle and sample image data of the sample wine bottle;

[0125] For each wine bottle, based on the feature information of the target bottle cap image of the wine bottle and the sample feature information of the sample wine bottle, a feature similarity recognition network is used to identify the feature difference information of the wine bottle;

[0126] identifying image difference information of the wine bottle based on the sample image data of the sample wine bottle and the target image data of the wine bottle;

[0127] The feature difference information of the wine bottle and the image difference information of the wine bottle are used as abnormal information of the target image data corresponding to the wine bottle.

[0128] Optionally, the identification module 330 is specifically configured to:

[0129] For each wine bottle, identifying the type of wine with a different cap based on the image difference information of the wine bottle;

[0130] Based on the characteristic difference information of the wine bottle, the probability of the wine with different caps in the target bottle cap image corresponding to the wine bottle is identified by using the probability identification strategy of the wine with different caps of the wine with different caps type.

[0131] Optionally, the identification module 330 is specifically configured to:

[0132] Obtaining a probability threshold value of wine with different caps for each type of wine with different caps;

[0133] Based on the probability threshold of each wine type with different caps and the probability of each wine type with different caps corresponding to each target bottle cap image, screening abnormal bottle cap images of each wine type with different caps in each target bottle cap image;

[0134] The type of wine with an abnormal cap corresponding to each abnormal bottle cap image and the probability of wine with an abnormal cap corresponding to each abnormal bottle cap image are used as information of each abnormal wine bottle.

[0135] Each module in the intelligent detection device for wine with a different cover can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in the computer device in the form of software, so that the processor can call and execute the corresponding operations of each module.

[0136] In an exemplary embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as shown in FIG. Figure 4 As shown. The computer device includes a processor, memory, an input / output interface, a communication interface, a display unit, and an input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are connected to the system bus via the input / output interface. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals via wired or wireless means, and the wireless means can be achieved via Wi-Fi, mobile cellular networks, NFC (near-field communication), or other technologies. When executed by the processor, the computer program implements an intelligent method for detecting wine with a foreign substance. The display unit of the computer device is used to produce a visually visible image and can be a display screen, a projection device, or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a button, trackball or touchpad set on the computer device casing, or an external keyboard, touchpad or mouse.

[0137] Those skilled in the art will understand that Figure 4 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0138] In an exemplary embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, steps corresponding to the intelligent detection method for wine with different caps are implemented.

[0139] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps corresponding to the intelligent detection method for wine with different caps are implemented.

[0140] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements steps corresponding to the intelligent detection method for wine with an abnormal cap.

[0141] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0142] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiments. In particular, any reference to memory, database, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The databases involved in the various embodiments provided herein may include at least one of a relational database and a non-relational database. Non-relational databases may include, but are not limited to, distributed databases based on blockchains. The processors involved in the various embodiments provided herein may be, but are not limited to, general-purpose processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), programmable logic devices (PLDs), data processing logic devices based on quantum computing, and the like.

[0143] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0144] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present application. It should be noted that a person of ordinary skill in the art may make various modifications and improvements without departing from the spirit of the present application, and these modifications and improvements fall within the scope of protection of the present application. Therefore, the scope of protection of the present application shall be determined by the appended claims.

Claims

1. An intelligent detection method for wine with different caps, characterized in that: The method comprises: Acquire a bottle cap image of each wine bottle, and perform image preprocessing on the bottle cap image of each wine bottle to obtain a target bottle cap image of each wine bottle; Based on the target bottle cap images of each of the wine bottles, feature information of each of the target bottle cap images is extracted through a feature extraction network, and based on the feature information of each of the target bottle cap images, abnormal information of each of the target bottle cap images is identified through a feature comparison strategy; Based on the abnormal information of each target bottle cap image, the abnormal wine evaluation strategy is used to identify the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image, and based on the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image, the information of each abnormal wine bottle is identified.

2. The method according to claim 1, characterized in that The performing image preprocessing on the bottle cap images of the wine bottles to obtain target bottle cap images of the wine bottles includes: For each wine bottle, split the bottle cap image into sub-bottle cap images corresponding to each photographing condition; By using the image optimization strategy corresponding to each of the photographing conditions, the sub-bottle cap images corresponding to each of the photographing conditions are optimized to obtain the target sub-bottle cap images corresponding to each of the photographing conditions; The target sub-bottle cap images corresponding to all photographing conditions are used as the target bottle cap image of the wine bottle.

3. The method according to claim 2, characterized in that The feature extraction network includes an image feature extraction network corresponding to each of the photographing conditions and a character feature extraction network corresponding to each of the photographing conditions; the feature extraction network extracts feature information of each of the target bottle cap images based on the target bottle cap images of each of the wine bottles, including: For each wine bottle, based on the target sub-bottle cap image corresponding to each of the photographing conditions, the image feature data corresponding to each of the photographing conditions is extracted through the image feature extraction network of each of the photographing conditions; Based on the target sub-bottle cap image corresponding to each of the photographing conditions, extracting the character feature data corresponding to each of the photographing conditions through the character feature extraction network of each of the photographing conditions; The image feature data corresponding to each of the photographing conditions and the character feature data corresponding to each of the photographing conditions are used as feature information of the target bottle cap image.

4. The method according to claim 1, wherein The identifying abnormal information of each target bottle cap image by a feature comparison strategy based on the feature information of each target bottle cap image includes: Acquiring sample feature information of a sample wine bottle and sample image data of the sample wine bottle; For each wine bottle, based on the feature information of the target bottle cap image of the wine bottle and the sample feature information of the sample wine bottle, a feature similarity recognition network is used to identify the feature difference information of the wine bottle; identifying image difference information of the wine bottle based on the sample image data of the sample wine bottle and the target image data of the wine bottle; The feature difference information of the wine bottle and the image difference information of the wine bottle are used as abnormal information of the target image data corresponding to the wine bottle.

5. The method according to claim 4, characterized in that The method of identifying the probability of an alien-cap wine of an alien-cap wine type corresponding to each target bottle cap image based on the abnormal information of each target bottle cap image by using an alien-cap wine evaluation strategy includes: For each wine bottle, identifying the type of wine with a different cap based on the image difference information of the wine bottle; Based on the characteristic difference information of the wine bottle, the probability of the target bottle cap image corresponding to the wine bottle is identified by the probability identification strategy of the wine with different caps of the wine with different caps type.

6. The method according to claim 1, characterized in that The method of identifying each abnormal wine bottle information based on the probability of the wine type with an abnormal cap corresponding to each target bottle cap image includes: Obtaining a probability threshold value of wine with different caps for each type of wine with different caps; Based on the probability threshold of each wine type with different caps and the probability of each wine type with different caps corresponding to each target bottle cap image, screening abnormal bottle cap images of each wine type with different caps in each target bottle cap image; The type of wine with an abnormal cap corresponding to each abnormal bottle cap image and the probability of wine with an abnormal cap corresponding to each abnormal bottle cap image are used as information of each abnormal wine bottle.

7. An intelligent detection device for wine with different caps, characterized in that: The device comprises: an acquisition module, configured to acquire the bottle cap image of each wine bottle and perform image preprocessing on the bottle cap image of each wine bottle to obtain a target bottle cap image of each wine bottle; an extraction module for extracting feature information of each target bottle cap image based on the target bottle cap image of each wine bottle through a feature extraction network, and identifying abnormal information of each target bottle cap image through a feature comparison strategy based on the feature information of each target bottle cap image; The recognition module is used to identify the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image based on the abnormal information of each target bottle cap image through the abnormal wine evaluation strategy, and identify the information of each abnormal wine bottle based on the abnormal wine probability of the abnormal wine type corresponding to each target bottle cap image.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images