A processing method and system for recognizing and positioning a plurality of illumination fusion pads

Abstract

The application provides a processing method and system for recognizing and positioning a plurality of illumination fusion pads, wherein the method comprises the following steps: step 1, providing an infrared radiation source at the bottom of a pad product; step 2, acquiring an infrared image and a visible light image of the pad product respectively; step 3, performing image fusion processing on the infrared image and the visible light image to obtain a fusion image; and step 4, recognizing and positioning the area of the pad of the pad product in the fusion image based on a template matching technology. The processing method and system for recognizing and positioning a plurality of illumination fusion pads can replace manual recognition of the welding position on the pad, realize fully automated welding, and greatly reduce the labor cost.

Description

Technical Field

[0001] This invention relates to the field of image recognition technology, and in particular to a processing method and system for recognizing and locating pads using multiple lighting fusion. Background Technology

[0002] Currently, when welding machines (such as tin soldering machines and laser welding machines) weld pads, manual identification of welding positions (such as solder joints and solder holes) is required, and the information is input into the welding machine one by one. The welding machine then performs welding operations based on the manually input welding positions, which results in high labor costs and makes it impossible to achieve fully automated welding.

[0003] Therefore, a solution is urgently needed. Summary of the Invention

[0004] One of the objectives of this invention is to provide a method and system for processing and locating solder pads by integrating multiple lighting, which replaces manual identification of the soldering position on the solder pads, realizes fully automated soldering, and greatly reduces labor costs.

[0005] This invention provides a method for processing multiple lighting fusion identification and positioning pads, comprising:

[0006] Step 1: Provide an infrared radiation source on the bottom of the product with solder pads;

[0007] Step 2: Acquire infrared and visible light images of the front of the product, respectively;

[0008] Step 3: Perform image fusion processing on the infrared image and the visible light image to obtain a fused image;

[0009] Step 4: Based on template matching technology, identify and locate the solder pads on the product according to the fused image.

[0010] Preferably, the infrared radiation source includes one or more combinations of: a glass-encapsulated wavelength filter heating filament, an infrared LED or LED array, and an infrared laser and lens.

[0011] Preferably, step 2: acquiring infrared and visible light images of the front of the product, respectively, including:

[0012] Acquire an infrared image of the front of the product using an infrared camera;

[0013] A visible light image of the front of the product is obtained using a CCD camera.

[0014] Preferably, step 3: performing image fusion processing on the infrared image and the visible light image to obtain a fused image, including:

[0015] Based on preset fusion rules, infrared and visible light images are fused to obtain a fused image.

[0016] Preferably, step 4: Based on template matching technology, the solder pads on the product are identified and located according to the fused image, including:

[0017] Obtain a preset pad recognition template library, which includes: multiple sets of one-to-one corresponding first template images and first recognition classification results;

[0018] The fused image is matched with the first template image;

[0019] If a match is found, the matching image block in the fused image is identified, and the first recognition and classification result corresponding to the first matching template image is marked next to the image block.

[0020] Preferably, the processing method for identifying and locating pads by fusion of multiple lighting also includes:

[0021] The pad recognition template library is expanded at preset time intervals;

[0022] This includes expanding the pad recognition template library, including:

[0023] Retrieve multiple supplementary contents;

[0024] Iterate through and insert the content in sequence;

[0025] During each iteration, the second template image and the second recognition and classification result are extracted from the inserted content.

[0026] Based on the preset first feature extraction template, feature extraction is performed on the second template image to obtain multiple first feature values;

[0027] Based on the first feature value, construct the first image feature description vector of the second template image;

[0028] Obtain historical matching data of the first template image corresponding to the first recognition classification result that is the same as the second recognition classification result in the pad recognition template library;

[0029] Based on the preset second feature extraction template, feature extraction is performed on historical matching cases to obtain multiple second feature values;

[0030] Based on the second feature value, construct a first case description vector for historical matching cases;

[0031] Retrieve a pre-defined database of matching compliance evaluation criteria;

[0032] Determine the first evaluation value corresponding to the first case description vector from the matching condition evaluation library;

[0033] If the evaluation value is greater than or equal to the preset first evaluation value threshold, feature extraction is performed on the corresponding first template image based on the first feature extraction template to obtain multiple third feature values;

[0034] Based on the third feature value, construct the second image feature description vector of the corresponding first template image;

[0035] Calculate the vector similarity between the first image feature description vector and any second image feature description vector;

[0036] If the vector similarity is greater than or equal to the preset vector similarity threshold, the traversed input content will be added to the pad recognition template library.

[0037] Preferably, multiple supplementary contents are obtained, including:

[0038] Retrieve multiple first pre-filled contents from local storage;

[0039] Gain experience points from the publisher of the first pre-filled content;

[0040] If the experience value is greater than or equal to the preset experience value threshold, the corresponding first pre-filled content will be used as the filled content;

[0041] And / or,

[0042] Multiple second pre-filled contents were obtained from the big data platform;

[0043] Obtain the guarantee from the big data platform regarding the credibility of the second pre-added content;

[0044] Based on the preset third feature extraction template, feature extraction is performed on the guarantee situation to obtain multiple fourth feature values;

[0045] Based on the fourth feature value, construct a second-case description vector for the guarantee situation;

[0046] Obtain the pre-defined guarantee status evaluation database;

[0047] Determine the second evaluation value corresponding to the second situation description vector from the guarantee situation evaluation database;

[0048] If the second evaluation value is greater than or equal to the preset second evaluation value threshold, the corresponding second pre-filled content will be used as the filled content.

[0049] This invention provides a processing system for identifying and locating pads using multiple lighting fusion methods, comprising:

[0050] Provide a module for providing an infrared radiation source on the bottom of a product with solder pads;

[0051] The acquisition module is used to acquire infrared and visible light images of the front of the product, respectively.

[0052] The processing module is used to perform image fusion processing on infrared and visible light images to obtain a fused image;

[0053] The positioning module is used to identify and locate the pads on the product based on template matching technology and the fused image.

[0054] Preferably, the infrared radiation source includes one or more combinations of: a glass-encapsulated wavelength filter heating filament, an infrared LED or LED array, and an infrared laser and lens.

[0055] Preferably, the acquisition module acquires infrared and visible light images of the front of the product, respectively, including:

[0056] Acquire an infrared image of the front of the product using an infrared camera;

[0057] A visible light image of the front of the product is obtained using a CCD camera.

[0058] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0059] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0060] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0061] Figure 1 This is a schematic diagram of a method for processing multiple lighting fusion identification and positioning pads in an embodiment of the present invention;

[0062] Figure 2 This is a schematic diagram of the preset fusion rules in an embodiment of the present invention;

[0063] Figure 3 This refers to the inverted and enhanced infrared image in this embodiment of the invention.

[0064] Figure 4 The image is a CCD image from an embodiment of the present invention.

[0065] Figure 5 This is the weighting graph in the embodiments of the present invention;

[0066] Figure 6This is the final fused image obtained in the embodiments of the present invention;

[0067] Figure 7 This is a schematic diagram of a multi-lighting fusion identification and positioning pad processing system according to an embodiment of the present invention. Detailed Implementation

[0068] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0069] This invention provides a method for processing multiple lighting fusion identification and positioning pads, such as... Figure 1 As shown, it includes:

[0070] Step 1: Provide an infrared radiation source on the bottom of the product with solder pads;

[0071] Step 2: Acquire infrared and visible light images of the front of the product, respectively;

[0072] Step 3: Perform image fusion processing on the infrared image and the visible light image to obtain a fused image;

[0073] Step 4: Based on template matching technology, identify and locate the solder pads on the product according to the fused image.

[0074] The working principle and beneficial effects of the above technical solution are as follows:

[0075] Infrared radiation sources provide better infrared radiation for products with solder pads. Infrared and visible light images of the product's front are acquired separately and then fused. Based on template matching technology, the solder pads on the product are identified and located according to the fused image. This can be applied to automatic soldering machines (including laser soldering) to automatically identify the position of solder pads and complete automatic soldering. It is particularly effective for solder pads on FPCs after tinning, significantly improving recognition accuracy. This is because the height, width, length, and even shape of the solder pads change after tinning. During CCD imaging, the shadows of each solder pad vary considerably due to the height change, leading to low accuracy with commonly used template matching methods. This application uses image fusion processing of infrared and visible light images followed by template matching to improve recognition accuracy.

[0076] This application replaces manual identification of soldering positions on solder pads, achieving fully automated soldering and greatly reducing labor costs.

[0077] This invention provides a method for processing multiple lighting fusion identification and positioning pads. The infrared radiation source includes one or more of the following: an electric heating filament of a glass-encapsulated wavelength filter, an infrared LED or LED array, and an infrared laser and lens.

[0078] The working principle and beneficial effects of the above technical solution are as follows:

[0079] Infrared radiation sources can be one or more combinations of the following: a glass-encapsulated wavelength filter heating filament, an infrared LED or LED array, and an infrared laser and lens.

[0080] This invention provides a method for processing multiple illumination fusion identification and positioning pads. Step 2: Acquire infrared and visible light images of the front of the product, including:

[0081] Acquire an infrared image of the front of the product using an infrared camera;

[0082] A visible light image of the front of the product is obtained using a CCD camera.

[0083] The working principle and beneficial effects of the above technical solution are as follows:

[0084] Infrared images can be captured by infrared cameras, while visible light images can be captured by CCD cameras.

[0085] This invention provides a method for identifying and locating pads using multiple illumination fusion techniques. Step 3: Perform image fusion processing on infrared and visible light images to obtain a fused image, including:

[0086] Based on preset fusion rules, infrared and visible light images are fused to obtain a fused image.

[0087] The working principle and beneficial effects of the above technical solution are as follows:

[0088] The pre-defined fusion rules are as follows: (1) After inverting and enhancing the infrared image, the pre-fused image is obtained by using gradient constraints together with the visible light image; (2) The pre-fused image is decomposed using a low-rank decomposition image fusion algorithm (MDLATLRR, where the pre-fused image is decomposed by an image decomposition algorithm) to obtain the base layer; (3) The local structural similarity algorithm (local SSIM) is then used to obtain the similarity index A between the pre-fused image and the infrared image, and the similarity index B between the pre-fused image and the visible light image, as shown in the diagram. Figure 2As shown; (4) Only the pixels with the large similarity index are retained in the whole image (judged by the similarity index AB) to obtain a weight map; (5) The visible light detail layer obtained by decomposing the visible light image using MDLATLR is fused with the weight map; (6) The infrared light detail obtained by decomposing the infrared image using MDLATLR is summed with the weighted visible light detail obtained in the previous step using different coefficients to obtain the fused detail layer; (7) The base layer of step (2) is simply added to the fused detail layer of the previous step to generate the final fused image.

[0089] Figure 3 This is an infrared image after inversion and enhancement processing. Figure 4 This is a CCD image. Figure 5 The weighted graph obtained in (4). Figure 6 This is the final fused image.

[0090] This invention provides a method for identifying and locating pads using multiple lighting fusion techniques. Step 4: Based on template matching technology, the pads on the product are identified and located according to the fused image, including:

[0091] Obtain a preset pad recognition template library, which includes: multiple sets of one-to-one corresponding first template images and first recognition classification results;

[0092] The fused image is matched with the first template image;

[0093] If a match is found, the matching image block in the fused image is identified, and the first recognition and classification result corresponding to the first matching template image is marked next to the image block.

[0094] The working principle and beneficial effects of the above technical solution are as follows:

[0095] The multiple sets of one-to-one corresponding first template images and first recognition classification results are as follows: for example, the first template image is a weld hole image, and the first recognition classification result is a weld hole. The fused image is matched with the first template image. If the match is successful, the first recognition classification result corresponding to the matching first template image is marked next to the image block to achieve recognition and localization.

[0096] This invention provides a method for identifying and locating pads using multiple lighting fusion techniques. Step 4: Based on template matching technology, the pads on the product are identified and located according to the fused image, including:

[0097] Obtain a preset pad recognition template library, which includes: multiple sets of one-to-one corresponding first template images and first recognition classification results;

[0098] The fused image is matched with the first template image;

[0099] If a match is found, the matching image block in the fused image is identified, and the first recognition and classification result corresponding to the first matching template image is marked next to the image block.

[0100] The working principle and beneficial effects of the above technical solution are as follows:

[0101] The first template image and the first recognition classification result are specifically: a template image used to identify the type of each region on the solder pad and the corresponding recognition classification type, for example: a solder hole image and the type "solder hole". The fused image is matched with the first template image. During matching, the first template image is randomly shifted on the fused image, and then matched with the region image of the area to which the first template image has been shifted in the fused image. If a match is found, the region image of the area to which the first template image has been shifted is taken as the matching image block in the fused image. When a match exists, the first recognition classification result corresponding to the matching first template image is marked next to the image block. This improves the accuracy and efficiency of image recognition.

[0102] This invention provides a method for processing multiple lighting fusion identification and positioning pads, and further includes:

[0103] The pad recognition template library is expanded at preset time intervals;

[0104] This includes expanding the pad recognition template library, including:

[0105] Retrieve multiple supplementary contents;

[0106] Iterate through and insert the content in sequence;

[0107] During each iteration, the second template image and the second recognition and classification result are extracted from the inserted content.

[0108] Based on the preset first feature extraction template, feature extraction is performed on the second template image to obtain multiple first feature values;

[0109] Based on the first feature value, construct the first image feature description vector of the second template image;

[0110] Obtain historical matching data of the first template image corresponding to the first recognition classification result that is the same as the second recognition classification result in the pad recognition template library;

[0111] Based on the preset second feature extraction template, feature extraction is performed on historical matching cases to obtain multiple second feature values;

[0112] Based on the second feature value, construct a first case description vector for historical matching cases;

[0113] Retrieve a pre-defined database of matching compliance evaluation criteria;

[0114] Determine the first evaluation value corresponding to the first case description vector from the matching condition evaluation library;

[0115] If the evaluation value is greater than or equal to the preset first evaluation value threshold, feature extraction is performed on the corresponding first template image based on the first feature extraction template to obtain multiple third feature values;

[0116] Based on the third feature value, construct the second image feature description vector of the corresponding first template image;

[0117] Calculate the vector similarity between the first image feature description vector and any second image feature description vector;

[0118] If the vector similarity is greater than or equal to the preset vector similarity threshold, the traversed input content will be added to the pad recognition template library.

[0119] The working principle and beneficial effects of the above technical solution are as follows:

[0120] When identifying and locating solder pads, the process relies on a solder pad identification template library. The quality of this library directly determines the effectiveness of the identification and location process. Therefore, the solder pad identification template library needs to be expanded periodically.

[0121] The input content is obtained, which includes a second template image and a second recognition and classification result. Based on a preset first feature extraction template, features are extracted from the second template image to obtain multiple first feature values, specifically: image grayscale level, image sharpness, and image contrast, etc. Based on the first feature values, a first image feature description vector is constructed. The first feature extraction template is designed to fit the pre-defined template for extracting these second feature values.

[0122] Generally, the first template image that is frequently matched in the pad recognition template library has the following requirements for being used as a template image: for example, the gray level of the first template image that is frequently matched is close to the gray level of the fused image, which facilitates matching and confirmation.

[0123] Therefore, the historical matching data of the first template image corresponding to the first recognition classification result that is the same as the second recognition classification result is obtained from the pad recognition template library. The historical matching data specifically includes the number of times the first template image has been matched historically and the matching degree of each match. A preset second feature extraction template is introduced to extract features from the historical matching data, obtaining multiple second feature values. These second feature values ​​specifically include the number of times the image has been matched historically and the matching degree of each match. The second feature extraction template is a pre-defined template for extracting these second feature values. Based on the second feature values, a first case description vector is constructed. A preset matching data evaluation library is introduced, which contains the first evaluation value corresponding to the first case description vector. Generally, the more times the image has been matched historically and the greater the matching degree of each match, the more qualified the first template image is to be used as a template image, and the larger the first evaluation value corresponding to the constructed first case description vector. A first evaluation value corresponding to the first condition description vector is determined from the matching condition evaluation library. If the first evaluation value is greater than or equal to a preset first evaluation value threshold, it indicates that the corresponding first template image sufficiently meets the requirements for use as a template image. Based on the first feature extraction template, feature extraction is performed on the corresponding first template image to obtain multiple third feature values. The third feature values ​​specifically include: image grayscale level, image sharpness, and image contrast, etc. Based on the third feature values, a second image feature description vector is constructed.

[0124] Calculate the vector similarity between the first image feature description vector and any second image feature description vector. If the vector similarity is greater than or equal to the preset vector similarity threshold, it means that the second template image in the supplemented content is sufficient to meet the requirements for use as a template image. Then, add the corresponding supplemented content to the pad recognition template library to achieve expansion.

[0125] This application expands the pad recognition template library, fully ensuring the accuracy and efficiency of pad recognition and positioning in fused images using template matching. In addition, the supplementary content is screened to ensure that it meets the requirements for use as template images, which greatly improves the expansion quality of the pad recognition template library. Furthermore, the requirements for use as template images are determined based on the historical matching of the first template with the same recognition and classification results, thus improving applicability.

[0126] This invention provides a method for processing multiple lighting fusion identification and positioning pads, obtaining multiple supplementary contents, including:

[0127] Retrieve multiple first pre-filled contents from local storage;

[0128] Gain experience points from the publisher of the first pre-filled content;

[0129] If the experience value is greater than or equal to the preset experience value threshold, the corresponding first pre-filled content will be used as the filled content;

[0130] And / or,

[0131] Multiple second pre-filled contents were obtained from the big data platform;

[0132] Obtain the guarantee from the big data platform regarding the credibility of the second pre-added content;

[0133] Based on the preset third feature extraction template, feature extraction is performed on the guarantee situation to obtain multiple fourth feature values;

[0134] Based on the fourth feature value, construct a second-case description vector for the guarantee situation;

[0135] Obtain the pre-defined guarantee status evaluation database;

[0136] Determine the second evaluation value corresponding to the second situation description vector from the guarantee situation evaluation database;

[0137] If the second evaluation value is greater than or equal to the preset second evaluation value threshold, the corresponding second pre-filled content will be used as the filled content.

[0138] The working principle and beneficial effects of the above technical solution are as follows:

[0139] There are two ways to obtain the supplementary content:

[0140] The first method involves acquiring content locally. Local storage contains initial pre-added content published by different individuals (employees). These pre-added content includes template images and classification results. The individuals have pre-collected images of different regions on different pads. The publisher's experience value represents their ability and experience in collecting different types of images from various pads. If the experience value is greater than or equal to a preset threshold, the corresponding initial pre-added content is used as the supplementary content. This improves the quality of the supplementary content.

[0141] The second method involves acquiring data from a big data platform. This platform facilitates data sharing among different manufacturers using pad positioning and recognition. Secondary pre-filled content, including template images and classification results, is obtained from this platform. The platform guarantees the reliability of the secondary pre-filled content through guarantees of strength and duration. Based on a pre-defined third feature extraction template, features are extracted from the guarantee to obtain multiple fourth feature values, specifically guarantee strength and duration. The third feature extraction template is pre-defined to adapt to extracting these fourth feature values. A pre-defined guarantee evaluation library is introduced, containing second evaluation values ​​corresponding to the second situation description vector. Generally, a higher guarantee strength and longer guarantee duration indicate higher reliability of the corresponding secondary pre-filled content, resulting in a larger second evaluation value for the constructed second situation description vector. The second evaluation value corresponding to the second situation description vector is determined from the guarantee evaluation library. If the second evaluation value is greater than or equal to a pre-defined second evaluation value threshold, the corresponding secondary pre-filled content is used as the filler content, thus improving the quality of the filler content.

[0142] This invention provides a processing system for identifying and locating pads using multiple lighting fusion methods, such as... Figure 7 As shown, it includes:

[0143] Module 1 is provided for providing an infrared radiation source on the bottom of a product with solder pads;

[0144] Acquisition module 2 is used to acquire infrared and visible light images of the front of the product, respectively;

[0145] Processing module 3 is used to perform image fusion processing on infrared images and visible light images to obtain a fused image;

[0146] Positioning module 4 is used to identify and locate pads on the product based on template matching technology and the fused image.

[0147] This invention provides a processing system for multiple lighting fusion identification and positioning pads. The infrared radiation source includes one or more of the following: an electric heating filament of a glass-encapsulated wavelength filter, an infrared LED or LED array, and an infrared laser and lens.

[0148] This invention provides a processing system for identifying and locating pads using a combination of multiple lighting conditions. The acquisition module 2 acquires both infrared and visible light images of the front of the product, including:

[0149] Acquire an infrared image of the front of the product using an infrared camera;

[0150] A visible light image of the front of the product is obtained using a CCD camera.

[0151] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for processing multiple lighting fusion identification and positioning pads, characterized in that, include: Step 1: Provide an infrared radiation source on the bottom of the product with solder pads; Step 2: Acquire the infrared image and visible light image of the front of the product, respectively; Step 3: Perform image fusion processing on the infrared image and the visible light image to obtain a fused image; Step 4: Based on template matching technology, identify and locate the pads on the product according to the fused image; Step 3, which involves performing image fusion processing on the infrared image and the visible light image to obtain a fused image, includes: After inverting and enhancing the infrared image, a pre-fused image is obtained by combining it with the visible light image using gradient constraints. The pre-fused image is decomposed using the low-rank decomposition image fusion algorithm MDLATRR to obtain the base layer; The local structural similarity algorithm Local SSIM is used to obtain a similarity index A for the pre-fused image and the infrared image, and a similarity index B for the pre-fused image and the visible light image. The pixels in the entire image are retained only at positions with high similarity indices. A weighted map is obtained by judging the similarity index AB. The visible light detail layer obtained by decomposing the visible light image using MDLATLR is fused with the weight map to obtain a weighted visible light detail layer. The infrared detail layer obtained by decomposing the infrared image using MDLATLR is summed with the weighted visible detail layer using different coefficients to obtain a fused detail layer. The base layer and the fusion detail layer are added together to generate the final fused image; Step 4: Based on template matching technology, the pads on the product are identified and located according to the fused image, including: Obtain a preset pad recognition template library, which includes: multiple sets of one-to-one corresponding first template images and first recognition classification results; The fused image is matched with the first template image; If a match is found, the matching image block in the fused image is determined, and the first recognition and classification result corresponding to the matching first template image is marked next to the image block. Also includes: The pad recognition template library is expanded at preset time intervals; The expansion of the pad recognition template library includes: Retrieve multiple supplementary contents; Iterate through the inserted content in sequence; During each iteration, the second template image and the second recognition and classification result are extracted from the inserted content. Based on a preset first feature extraction template, feature extraction is performed on the second template image to obtain multiple first feature values; Based on the first feature value, a first image feature description vector of the second template image is constructed; Obtain the historical matching status of the first template image corresponding to the first recognition classification result that is the same as the second recognition classification result in the solder pad recognition template library; Based on a preset second feature extraction template, feature extraction is performed on the historical matching cases to obtain multiple second feature values; Based on the second feature value, a first case description vector of the historical matching cases is constructed; Retrieve a pre-defined database of matching compliance evaluation criteria; Determine the first evaluation value corresponding to the first situation description vector from the matching compliance evaluation library; If the evaluation value is greater than or equal to a preset first evaluation value threshold, feature extraction is performed on the corresponding first template image based on the first feature extraction template to obtain multiple third feature values; Based on the third feature value, a second image feature description vector corresponding to the first template image is constructed; Calculate the vector similarity between the first image feature description vector and any of the second image feature description vectors; If the vector similarity is greater than or equal to a preset vector similarity threshold, the traversed supplementary content will be added to the pad recognition template library.

2. The processing method for identifying and positioning pads by fusion of multiple lighting as described in claim 1, characterized in that, The infrared radiation source includes one or more of the following: an electrothermal filament of a glass-encapsulated wavelength filter, an infrared LED or LED array, and an infrared laser and lens.

3. The processing method for identifying and positioning pads by fusion of multiple lighting as described in claim 1, characterized in that, Step 2: Acquire infrared and visible light images of the front of the product, respectively, including: An infrared image of the front of the product is obtained using an infrared camera; A visible light image of the front of the product is acquired using a CCD camera.

4. The processing method for multi-lighting fusion identification and positioning pads as described in claim 1, characterized in that, Retrieve multiple supplementary contents, including: Retrieve multiple first pre-filled contents from local storage; Obtain the experience value of the publisher of the first pre-filled content; If the experience value is greater than or equal to the preset experience value threshold, the first pre-filled content will be used as the filled content; And / or, Multiple second pre-filled contents were obtained from the big data platform; Obtain the guarantee status of the big data platform regarding the credibility of the second pre-added content; Based on a preset third feature extraction template, feature extraction is performed on the guarantee situation to obtain multiple fourth feature values; Based on the fourth feature value, a second situation description vector for the guarantee situation is constructed; Obtain the pre-defined guarantee status evaluation database; Determine the second evaluation value corresponding to the second situation description vector from the guarantee situation evaluation database; If the second evaluation value is greater than or equal to the preset second evaluation value threshold, the corresponding second pre-filled content will be used as the filled content.

5. A processing system for identifying and locating pads by fusion of multiple lighting methods, characterized in that, include: Provide a module for providing an infrared radiation source on the bottom of a product with solder pads; The acquisition module is used to acquire infrared and visible light images of the front of the product, respectively. The processing module is used to perform image fusion processing on the infrared image and the visible light image to obtain a fused image; The positioning module is used to identify and locate the pads on the product based on the fused image using template matching technology. The process of fusing the infrared image and the visible light image to obtain a fused image includes: After inverting and enhancing the infrared image, a pre-fused image is obtained by combining it with the visible light image using gradient constraints. The pre-fused image is decomposed using the low-rank decomposition image fusion algorithm MDLATRR to obtain the base layer; The local structural similarity algorithm Local SSIM is used to obtain a similarity index A for the pre-fused image and the infrared image, and a similarity index B for the pre-fused image and the visible light image. The pixels in the entire image are retained only at positions with high similarity indices. A weighted map is obtained by judging the similarity index AB. The visible light detail layer obtained by decomposing the visible light image using MDLATLR is fused with the weight map to obtain a weighted visible light detail layer. The infrared detail layer obtained by decomposing the infrared image using MDLATLR is summed with the weighted visible detail layer using different coefficients to obtain a fused detail layer. The base layer and the fusion detail layer are added together to generate the final fused image; The method of identifying and locating solder pads on the product based on the template matching technology and the fused image includes: Obtain a preset pad recognition template library, which includes: multiple sets of one-to-one corresponding first template images and first recognition classification results; The fused image is matched with the first template image; If a match is found, the matching image block in the fused image is determined, and the first recognition and classification result corresponding to the matching first template image is marked next to the image block. It also includes expansion modules for: The pad recognition template library is expanded at preset time intervals; The expansion of the pad recognition template library includes: Retrieve multiple supplementary contents; Iterate through the inserted content in sequence; During each iteration, the second template image and the second recognition and classification result are extracted from the inserted content. Based on a preset first feature extraction template, feature extraction is performed on the second template image to obtain multiple first feature values; Based on the first feature value, a first image feature description vector of the second template image is constructed; Obtain the historical matching status of the first template image corresponding to the first recognition classification result that is the same as the second recognition classification result in the solder pad recognition template library; Based on a preset second feature extraction template, feature extraction is performed on the historical matching cases to obtain multiple second feature values; Based on the second feature value, a first case description vector of the historical matching cases is constructed; Retrieve a pre-defined database of matching compliance evaluation criteria; Determine the first evaluation value corresponding to the first situation description vector from the matching compliance evaluation library; If the evaluation value is greater than or equal to a preset first evaluation value threshold, feature extraction is performed on the corresponding first template image based on the first feature extraction template to obtain multiple third feature values; Based on the third feature value, a second image feature description vector corresponding to the first template image is constructed; Calculate the vector similarity between the first image feature description vector and any of the second image feature description vectors; If the vector similarity is greater than or equal to a preset vector similarity threshold, the traversed supplementary content will be added to the pad recognition template library.

6. The processing system for multi-lighting fusion identification and positioning pads as described in claim 5, characterized in that, The infrared radiation source includes one or more of the following: an electrothermal filament of a glass-encapsulated wavelength filter, an infrared LED or LED array, and an infrared laser and lens.

7. The processing system for multi-lighting fusion identification and positioning pads as described in claim 5, characterized in that, The acquisition module acquires infrared and visible light images of the front of the product, respectively, including: An infrared image of the front of the product is obtained using an infrared camera; A visible light image of the front of the product is acquired using a CCD camera.

Images