Adaptive distance measurement method and system for dangerous sources in substation based on binocular vision

Abstract

The present invention provides an adaptive ranging method and system for dangerous sources in a substation based on binocular vision, wherein the method comprises: step 1: controlling a binocular camera to obtain a visible light image in the substation; step 2: obtaining ranging environment conditions based on the visible light image; step 3: controlling an infrared thermal imager to obtain infrared thermal image data of different objects in the substation according to the ranging environment conditions; step 4: fusing and analyzing the visible light image and the infrared thermal image data to determine the ranging result of the dangerous source. The adaptive ranging method and system for dangerous sources in a substation based on binocular vision of the present invention obtains ranging environment conditions based on the visible light image in the substation obtained by the binocular camera. According to the ranging environment conditions, the infrared thermal imager is controlled to obtain infrared thermal image data of different objects in the substation, and the visible light image and the infrared thermal image data are fused and analyzed to determine the ranging result of the dangerous source, thereby improving the ranging accuracy and being safer.

Description

Technical Field

[0001] The present invention relates to the technical field of visual distance measurement, and in particular to a binocular vision-based adaptive distance measurement method and system for hazardous sources in a transformer substation. Background Art

[0002] Substations usually contain high-voltage equipment and complex electrical systems. Accurately identifying and evaluating hazardous sources is essential to preventing accidents and ensuring personnel safety. At present, binocular vision adaptive ranging technology plays an important role in substation safety management. It uses two cameras (simulating the two eyes of a person) to capture images of the same scene from different angles, and uses the parallax between the two images to calculate the distance of the object, providing accurate spatial information, helping to prevent accidents and ensure the safety of equipment and personnel.

[0003] The invention patent with application number: CN202210689777.8 discloses a method, system, medium and equipment for automatic inspection of substations, wherein the method includes: by means of identification matching, the inspection robot can change the inspection route at any time as needed. The position of the robot and the position of the abnormal condition are determined by combining binocular vision technology with three-dimensional spatial positioning technology. High-precision and high-quality inspection work in the substation is achieved, abnormalities in the substation can be discovered in time, and maintenance personnel can be notified to rush to the scene for fault handling as soon as possible.

[0004] However, when the above-mentioned prior art performs positioning through binocular vision, when the visible light conditions are not suitable (for example, the brightness is low, there are reflective lights on the surface of the object), the positioning information obtained from the binocular camera is also inaccurate.

[0005] In view of this, there is an urgent need for an adaptive distance measurement method and system for hazardous sources in substations based on binocular vision to at least solve the above-mentioned deficiencies. Summary of the invention

[0006] One of the purposes of the present invention is to provide a binocular vision-based adaptive distance measurement method and system for dangerous sources in a substation, and obtain the distance measurement environment based on the visible light image in the substation obtained by the binocular camera. When it is determined that infrared thermal imaging is required, the infrared thermal imager is controlled to obtain infrared thermal image data of different objects in the substation, and the visible light image and the infrared thermal image data are fused and analyzed to determine the distance measurement result of the dangerous source, thereby avoiding the problem of inaccurate distance measurement results of dangerous sources determined only based on binocular camera images when the light conditions are not suitable, and being safer.

[0007] The embodiment of the present invention provides a binocular vision-based adaptive distance measurement method for hazardous sources in a substation, comprising:

[0008] Step 1: Control the binocular camera to obtain visible light images in the substation;

[0009] Step 2: Based on the visible light image, obtain the ranging environment;

[0010] Step 3: When the ranging environment requires infrared thermal imaging, the infrared thermal imager is controlled to obtain infrared thermal image data of different objects in the substation;

[0011] Step 4: Fuse and analyze the visible light image and infrared thermal image data, and determine the ranging results of the hazard source.

[0012] Preferably, step 1: controlling a binocular camera to acquire a visible light image in a substation includes:

[0013] Get the station map inside the substation;

[0014] Plan the binocular camera image acquisition trajectory based on the station map;

[0015] According to the binocular camera image acquisition trajectory, the binocular camera is controlled to obtain the visible light image in the substation.

[0016] The binocular vision-based adaptive distance measurement method for hazardous sources in a substation provided by an embodiment of the present invention further includes:

[0017] When controlling the binocular camera to obtain the visible light image in the substation, multiple binocular camera signals are registered to obtain the visible light image;

[0018] Among them, the multi-channel binocular camera signals are registered to obtain visible light images, including:

[0019] During registration, the spatial correlation between the reference image and the local area image is obtained. The calculation formula of the spatial correlation is as follows:

[0020]

[0021] Wherein, T(p,q) is the spatial correlation of the local area image local relative to the reference image St after being offset by p in the horizontal direction and q in the vertical direction, local(a,b) is the pixel value of the local area image local at position (a,b), and St(ap,bq) is the pixel value of the reference image St at position (ap,bq);

[0022] When the spatial correlation reaches the maximum value, the registration is completed and the visible light image is obtained.

[0023] Preferably, step 2: obtaining ranging environment conditions based on the visible light image includes:

[0024] Based on the visible light image, the lighting condition analysis is performed to obtain lighting analysis items; the analysis items include: image brightness, image contrast, whether the image lighting is uniform, and image exposure;

[0025] Obtain standard ranging environment condition determination factors;

[0026] According to the illumination analysis item, determine the current ranging environment condition determination factor;

[0027] If the current ranging environment determination factor matches the standard ranging environment determination factor, then the ranging environment does not require infrared thermal imaging;

[0028] If there is no match between the current ranging environment determination factor and the standard ranging environment determination factor, the ranging environment requires infrared thermal imaging.

[0029] Preferably, obtaining the standard ranging environment condition determination factor includes:

[0030] Obtain a target ranging record set; wherein the target ranging record set includes multiple target ranging records;

[0031] A control group of target ranging records is obtained; wherein the control group is a ranging record that integrates binocular camera images and infrared sensor data in the target ranging records;

[0032] Determine the difference in distance measurement results based on the target distance measurement records and the control group;

[0033] Determine the illumination conditions of the target ranging records whose ranging result difference is less than or equal to a preset ranging result difference threshold, and use them as extraction conditions;

[0034] A template is constructed based on the extraction conditions and factors, and the factors for determining the standard ranging environment are constructed.

[0035] Preferably, obtaining a target ranging record set includes:

[0036] Obtaining the first pre-selected target ranging record;

[0037] Parsing the first pre-selected target ranging record, and obtaining parameter similarity between binocular camera parameters of a binocular camera used in the first pre-selected target ranging record and binocular camera parameters of a binocular camera for shooting a visible light image;

[0038] If the parameter similarity is greater than or equal to a preset parameter similarity threshold, the first pre-selected target ranging record is used as the second pre-selected target ranging record;

[0039] Traversing each second pre-selected target ranging record in turn, and obtaining condition similarities between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to other second pre-selected target ranging records;

[0040] If the conditional similarities are all less than or equal to the preset conditional similarity threshold, the second pre-selected target ranging record being traversed is used as the target ranging record;

[0041] If there is a condition similarity greater than a preset condition similarity threshold, stop calculating the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records;

[0042] Determine whether the calculation record with conditional similarity greater than a preset conditional similarity threshold is calculated for the first time, if so, take the second pre-selected target ranging record being traversed as the target ranging record and continue to traverse the second pre-selected target ranging record that has not been traversed, if not, directly continue to traverse the second pre-selected target ranging record that has not been traversed;

[0043] After the traversal is completed, all target distance measurement records are integrated to obtain a target distance measurement record set.

[0044] The binocular vision-based adaptive distance measurement method for hazardous sources in a substation provided by an embodiment of the present invention further includes:

[0045] Before stopping calculating the condition similarity between the illumination condition corresponding to the second pre-selected target ranging record being traversed and the illumination condition corresponding to the remaining second pre-selected target ranging records, determining whether the shooting angles of the second pre-selected target ranging record being traversed and the second pre-selected target ranging record corresponding to the calculated condition similarity are consistent;

[0046] If they are consistent, stop calculating the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records, and continue traversing the second pre-selected target ranging records that have not been traversed;

[0047] If they are inconsistent, continue to calculate the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records;

[0048] If the condition similarity between the lighting condition corresponding to the second pre-selected target ranging record being traversed and the lighting condition corresponding to the last second pre-selected target ranging record for condition similarity calculation is less than or equal to a preset condition similarity threshold, or the condition similarity is greater than the preset condition similarity threshold and the corresponding shooting angles are inconsistent, the second pre-selected target ranging record being traversed is used as the target ranging record.

[0049] The embodiment of the present invention provides a binocular vision-based adaptive distance measurement system for hazardous sources in a substation, comprising:

[0050] Binocular camera control subsystem, used to control the binocular camera to obtain visible light images in the substation;

[0051] A ranging environment acquisition subsystem is used to acquire ranging environment conditions based on visible light images;

[0052] The infrared thermal imager control subsystem is used to control the infrared thermal imager to obtain infrared thermal image data of different objects in the substation when the ranging environment requires infrared thermal imaging;

[0053] The hazard source ranging subsystem is used to fuse and analyze visible light images and infrared thermal imaging data and determine the ranging results of the hazard source.

[0054] Preferably, the binocular camera control subsystem includes:

[0055] The station map acquisition module is used to obtain the station map of the substation;

[0056] The acquisition trajectory planning module is used to plan the binocular camera image acquisition trajectory according to the station map;

[0057] The visible light image acquisition module is used to control the binocular camera to acquire the visible light image in the substation according to the binocular camera image acquisition trajectory.

[0058] The binocular vision-based adaptive distance measurement system for hazardous sources in a substation provided by an embodiment of the present invention further includes:

[0059] The signal registration module is used to control the binocular camera to obtain the visible light image in the substation, and to register the signals of multiple binocular cameras to obtain the visible light image;

[0060] Among them, the signal registration module performs the following operations:

[0061] During registration, the spatial correlation between the reference image and the local area image is obtained. The calculation formula of the spatial correlation is as follows:

[0062]

[0063]

[0064] Wherein, T(p,q) is the spatial correlation of the local area image local relative to the reference image St after being offset by p in the horizontal direction and q in the vertical direction, local(a,b) is the pixel value of the local area image local at position (a,b), and St(ap,bq) is the pixel value of the reference image St at position (ap,bq);

[0065] When the spatial correlation reaches the maximum value, the registration is completed and the visible light image is obtained.

[0066] The beneficial effects of the present invention are:

[0067] The present invention obtains the ranging environment based on the visible light image in the substation obtained by the binocular camera. When it is determined that infrared thermal imaging is needed, the infrared thermal imager is controlled to obtain infrared thermal image data of different objects in the substation, and the visible light image and the infrared thermal image data are fused and analyzed to determine the ranging result of the dangerous source, thereby avoiding the problem of inaccurate ranging results of dangerous sources determined only based on binocular camera images when the light conditions are not suitable, and is safer.

[0068] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in the present application documents.

[0069] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0071] Figure 1 Schematic diagram of an adaptive distance measurement method for hazardous sources in a substation based on binocular vision in an embodiment of the present invention;

[0072] Figure 2 Schematic diagram of an adaptive distance measurement system for hazardous sources in a substation based on binocular vision in an embodiment of the present invention. DETAILED DESCRIPTION

[0073] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0074] The embodiment of the present invention provides a method for adaptively measuring the distance between dangerous sources in a substation based on binocular vision, such as Figure 1 As shown, including:

[0075] Step 1: Control the binocular camera to obtain the visible light image in the substation; wherein the visible light image is: the image in the substation taken by the binocular camera, the binocular camera can be set at a fixed point in the substation, can also be set on an automatic cruising robot, or can be controlled by a handheld binocular camera; when controlling the binocular camera to shoot, the handheld binocular camera is manually shot autonomously, and the binocular camera at a fixed point or set on the robot can shoot according to a preset shooting strategy; the shooting strategy is, for example: detecting the danger source based on the target detection technology, if the detection is successful, then shooting, and for example: shooting based on a preset shooting cycle;

[0076] Step 2: Based on the visible light image, the ranging environment is obtained; wherein the ranging environment is divided into: requiring infrared thermal imaging and not requiring infrared thermal imaging; when obtaining the ranging environment based on the visible light image, it is determined whether the subsequent hazard source ranging can be accurately performed based on the visible light image, if so, infrared thermal imaging is not required, otherwise infrared thermal imaging is required;

[0077] Step 3: When the ranging environment requires infrared thermal imaging, the infrared thermal imager is controlled to obtain infrared thermal image data of different objects in the substation; wherein the infrared thermal image data is: image data captured by the infrared thermal imager, such as: infrared thermal image data of the transformer;

[0078] Step 4: Fuse and analyze the visible light image and infrared thermal image data, and determine the ranging results of the hazard source.

[0079] The working principle and beneficial effects of the above technical solution are:

[0080] This application obtains the ranging environment based on the visible light image of the substation obtained by the binocular camera. When it is determined that infrared thermal imaging is needed, the infrared thermal imager is controlled to obtain infrared thermal image data of different objects in the substation, and the visible light image and infrared thermal image data are fused and analyzed to determine the ranging result of the dangerous source, avoiding the problem of inaccurate ranging results of dangerous sources determined only based on binocular camera images when the light conditions are not suitable, which is safer.

[0081] In one embodiment, step 1: controlling a binocular camera to acquire a visible light image in a substation includes:

[0082] Obtaining a station map within the substation; wherein the station map is: a distribution map of equipment within the substation;

[0083] According to the station map, plan the binocular camera image acquisition trajectory; the image acquisition trajectory is: the acquisition path of the binocular camera to collect visible light images;

[0084] According to the binocular camera image acquisition trajectory, the binocular camera is controlled to obtain the visible light image in the substation.

[0085] The working principle and beneficial effects of the above technical solution are:

[0086] This application introduces an in-station map within the substation, and plans a binocular camera image acquisition trajectory for the binocular camera to acquire visible light images within the in-station map. Based on the binocular camera image acquisition trajectory, the binocular camera is controlled to acquire visible light images within the substation, thereby improving the efficiency of acquiring visible light images.

[0087] The embodiment of the present invention provides a binocular vision-based adaptive distance measurement method for hazardous sources in a substation, further comprising:

[0088] When controlling the binocular camera to obtain the visible light image in the substation, multiple binocular camera signals are registered to obtain the visible light image;

[0089] Among them, the multi-channel binocular camera signals are registered to obtain visible light images, including:

[0090] During registration, the spatial correlation between the reference image and the local area image is obtained. The calculation formula of the spatial correlation is as follows:

[0091]

[0092] Wherein, T(p,q) is the spatial correlation of the local area image local relative to the reference image St after being offset by p in the horizontal direction and q in the vertical direction, local(a,b) is the pixel value of the local area image local at position (a,b), and St(ap,bq) is the pixel value of the reference image St at position (ap,bq);

[0093] When the spatial correlation reaches the maximum value, the registration is completed and the visible light image is obtained.

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

[0095] The present application obtains visible light images by registering local area images acquired by multiple binocular cameras, thereby improving the standardization of visible light image acquisition.

[0096] In one embodiment, step 2: obtaining ranging environment conditions based on the visible light image includes:

[0097] Based on the visible light image, perform illumination condition analysis to obtain illumination analysis items; the illumination analysis items include: image brightness, image contrast, image illumination uniformity, and image exposure; wherein the illumination condition analysis is to evaluate the quality and characteristics of illumination in the image;

[0098] Obtain the standard ranging environment condition determination factor; wherein the standard ranging environment condition determination factor is: a description vector of the lighting environment suitable for subsequent hazard source measurement, for example: a 1×4 vector, the first row and the first column represents the image brightness, the first row and the second column represents the image contrast, the first row and the third column represents whether the image lighting is uniform, and the first row and the fourth column represents the image exposure;

[0099] According to the illumination analysis item, determine the current ranging environment condition determination factor; wherein the construction rule of the current ranging environment condition determination factor is the same as the standard ranging environment condition determination factor, and the element value in the factor (vector) is determined by the illumination analysis item;

[0100] If the current ranging environment determination factor and the standard ranging environment determination factor match, the ranging environment does not require infrared thermal imaging; wherein, the matching means that the vector angle between the current ranging environment determination factor and any standard ranging environment determination factor is within a preset angle range, and the preset angle range is pre-set by the staff, such as: 0° to 5°;

[0101] If there is no match between the current ranging environment determination factor and the standard ranging environment determination factor, the ranging environment requires infrared thermal imaging.

[0102] The working principle and beneficial effects of the above technical solution are:

[0103] It is inefficient to manually judge whether the lighting conditions of visible light images require the assistance of infrared thermal imaging. In addition, manual judgment also relies on the experience of the judge and is not reliable enough. Therefore, this application determines the lighting analysis items of the visible light image and obtains the standard ranging environment condition determination factor that describes the lighting environment suitable for subsequent hazard source measurement. The current ranging environment condition determination factor determined according to the lighting analysis item and the standard ranging environment condition determination factor are matched. If there is a match, infrared thermal imaging is not required. Otherwise, infrared thermal imaging is required, which improves the efficiency of determining lighting conditions.

[0104] In one embodiment, obtaining a standard ranging environment condition determination factor includes:

[0105] Obtain a target ranging record set; wherein the target ranging record set includes a plurality of target ranging records; the target ranging record is: an extraction record of the illumination condition used for constructing a standard ranging environment condition determination factor;

[0106] Obtaining a control group of target ranging records; wherein the control group is a ranging record that fuses the binocular camera image in the target ranging record and the infrared sensor data of the measured object in the target ranging record;

[0107] Determine the distance measurement result difference according to the target distance measurement record and the control group; wherein the distance measurement result difference is: the difference between the distance measurement result of the object measured by the target distance measurement record and the distance measurement result of the object measured by the control group, for example: the object A measured in the target distance measurement record is 22 meters, and the object A measured in the control group is 22.5 meters, then the distance measurement result difference is 0.5 meters;

[0108] Determine the illumination conditions of the target ranging records whose ranging result difference is less than or equal to a preset ranging result difference threshold, and use them as extraction conditions; wherein the preset ranging result difference threshold is manually preset;

[0109] Based on the extraction conditions and factor construction template, the standard ranging environment condition determination factor is constructed. Among them, the factor construction template is: a template for generating the standard ranging environment condition determination factor according to the extraction conditions, for example: the extraction conditions are image brightness B, image contrast C, image illumination uniformity and image exposure D, then the factor construction template inputs the corresponding factor value into the corresponding vector position to obtain the standard ranging environment condition determination factor, such as: [B, C, 1, D]. In this example, 1 represents uniform image illumination and 0 represents uneven image illumination. The specific vector element representation rules can be set manually.

[0110] The working principle and beneficial effects of the above technical solution are:

[0111] This application obtains a target ranging record set constructed by recording the process of ranging based on the images taken by the binocular camera in history. At the same time, a control group of the target ranging records is obtained to determine the difference in ranging results between the object ranging results measured by the target ranging records and the object ranging results measured by the control group. If the difference in ranging results is less than or equal to the preset ranging result difference threshold, it means that the lighting conditions in the corresponding target ranging records have little effect on the final ranging results and meet the extraction requirements of the standard lighting conditions, so it is used as the extraction condition. A factor construction template is introduced, and the factor construction template is used for the extraction conditions to generate the standard ranging environment condition determination factor, and the determination process of the standard ranging environment condition determination factor is more reasonable.

[0112] In one embodiment, obtaining a target ranging record set includes:

[0113] Acquire a first pre-selected target ranging record; wherein the first pre-selected target ranging record is: a ranging process record based on images captured by a binocular camera in history;

[0114] Parse the first pre-selected target ranging record to obtain the parameter similarity between the binocular camera parameters of the binocular camera used in the first pre-selected target ranging record and the binocular camera parameters of the binocular camera for shooting the visible light image; wherein the binocular camera parameters are: device parameters of the binocular camera, such as: resolution, field of view angle, focal length, etc.; parameter similarity is: parameter similarity degree;

[0115] If the parameter similarity is greater than or equal to a preset parameter similarity threshold, the first pre-selected target distance measurement record is used as the second pre-selected target distance measurement record; wherein the preset parameter similarity threshold is manually preset;

[0116] Traversing each second pre-selected target ranging record in turn, obtaining the condition similarity between the illumination condition corresponding to the second pre-selected target ranging record being traversed and the illumination condition corresponding to other second pre-selected target ranging records; wherein the condition similarity is: the similarity degree of illumination conditions;

[0117] If the conditional similarities are all less than or equal to the preset conditional similarity threshold, the second pre-selected target ranging record being traversed is used as the target ranging record;

[0118] If there is a condition similarity greater than a preset condition similarity threshold, stop calculating the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records;

[0119] Determine whether the calculation record with conditional similarity greater than a preset conditional similarity threshold is calculated for the first time. If so, take the second pre-selected target ranging record being traversed as the target ranging record and continue to traverse the second pre-selected target ranging record that has not been traversed yet. If not, directly continue to traverse the second pre-selected target ranging record that has not been traversed yet; wherein the calculation record is: the second pre-selected target ranging record currently being traversed and the conditional similarity greater than the preset conditional similarity threshold and other second pre-selected target ranging records; the first calculation is: the record subjects of the calculation record are all the first time that the conditional similarity calculation result of the corresponding illumination condition meets the conditional similarity greater than the preset conditional similarity threshold;

[0120] After the traversal is completed, all target distance measurement records are integrated to obtain a target distance measurement record set.

[0121] The working principle and beneficial effects of the above technical solution are:

[0122] When obtaining the target distance measurement record set, it can be obtained from historical data. However, not all historical data are suitable for subsequent standard illumination condition information extraction. For example, the parameters of the binocular camera are different, and its sensitivity to changes in illumination conditions is also different. Therefore, it is necessary to screen the first pre-selected target distance measurement records in history. When screening, calculate the parameter similarity between the binocular camera parameters of the binocular camera used in the first pre-selected target distance measurement record and the binocular camera parameters of the binocular camera that shoots the visible light image, and screen out the second pre-selected target distance measurement record whose parameter similarity is greater than or equal to the preset parameter similarity threshold.

[0123] In addition, when constructing the target ranging record set, it is necessary to consider obtaining as comprehensive standard lighting condition information as possible. Therefore, the condition similarity between the lighting conditions corresponding to the second pre-selected target ranging record being traversed and the lighting conditions corresponding to other second pre-selected target ranging records is calculated. If the condition similarity is less than or equal to the preset condition similarity threshold, it means that the provided lighting condition information is comprehensive and rich and should be adopted. If the condition similarity is greater than the preset condition similarity threshold, it means that the provided lighting condition information is similar. In order to avoid data redundancy, it is appropriately eliminated, and the traversal of the next second pre-selected target ranging record is continued to select the target ranging record. When all the second pre-selected target ranging records are traversed, all target ranging records are summarized to obtain the target ranging record set, which improves the acquisition efficiency of the target ranging record set.

[0124] The embodiment of the present invention provides a binocular vision-based adaptive distance measurement method for hazardous sources in a substation, further comprising:

[0125] Before stopping calculating the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records, determining whether the shooting angles of the second pre-selected target ranging record being traversed and the corresponding second pre-selected target ranging record for which the condition similarity is calculated are consistent; wherein the shooting angle is: the shooting angle of the binocular camera shooting the object in the second pre-selected target ranging record;

[0126] If they are consistent, stop calculating the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records, and continue traversing the second pre-selected target ranging records that have not been traversed;

[0127] If they are inconsistent, continue to calculate the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records;

[0128] If the condition similarity between the lighting condition corresponding to the second pre-selected target ranging record being traversed and the lighting condition corresponding to the last second pre-selected target ranging record for condition similarity calculation is less than or equal to a preset condition similarity threshold, or the condition similarity is greater than the preset condition similarity threshold and the corresponding shooting angles are inconsistent, the second pre-selected target ranging record being traversed is used as the target ranging record.

[0129] The working principle and beneficial effects of the above technical solution are:

[0130] Generally, when the lighting conditions are similar under the same viewing angle, there is information redundancy when repeatedly extracting the lighting conditions to construct the standard ranging environment condition determination factor. However, different shooting viewing angles have different requirements for lighting conditions. When the lighting conditions are similar, they may also be under different viewing angles. At this time, it is unreasonable to blindly eliminate the illumination information. Therefore, if there is a conditional similarity greater than the preset conditional similarity threshold, before stopping the calculation of the conditional similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records, it is determined whether the shooting angles of the second pre-selected target ranging record being traversed and the corresponding second pre-selected target ranging record for calculating the conditional similarity are consistent. If they are consistent, the second pre-selected target ranging record being traversed is not considered as the target ranging record. If they are inconsistent, the calculation continues. If the conditional similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the last second pre-selected target ranging record for conditional similarity calculation is less than or equal to the preset conditional similarity threshold, or the conditional similarity is greater than the preset conditional similarity threshold and the corresponding shooting angles are inconsistent, the second pre-selected target ranging record being traversed is used as the target ranging record. The suitability of the target ranging record screening is improved.

[0131] The embodiment of the present invention provides an adaptive distance measurement system for dangerous sources in a substation based on binocular vision, such as Figure 2 As shown, including:

[0132] The binocular camera control subsystem 1 is used to control the binocular camera to obtain visible light images in the substation;

[0133] The ranging environment condition acquisition subsystem 2 is used to acquire the ranging environment condition based on the visible light image;

[0134] The infrared thermal imager control subsystem 3 is used to control the infrared thermal imager to obtain infrared thermal image data of different objects in the substation when the ranging environment requires infrared thermal imaging;

[0135] The hazard source distance measurement subsystem 4 is used to fuse and analyze the visible light image and infrared thermal image data and determine the distance measurement result of the hazard source.

[0136] In one embodiment, the binocular camera control subsystem includes:

[0137] The station map acquisition module is used to obtain the station map of the substation;

[0138] The acquisition trajectory planning module is used to plan the binocular camera image acquisition trajectory according to the station map;

[0139] The visible light image acquisition module is used to control the binocular camera to acquire the visible light image in the substation according to the binocular camera image acquisition trajectory.

[0140] The embodiment of the present invention provides a binocular vision-based adaptive distance measurement system for hazardous sources in a substation, further comprising:

[0141] The signal registration module is used to control the binocular camera to obtain the visible light image in the substation, and to register the signals of multiple binocular cameras to obtain the visible light image;

[0142] Among them, the signal registration module performs the following operations:

[0143] During registration, the spatial correlation between the reference image and the local area image is obtained. The calculation formula of the spatial correlation is as follows:

[0144]

[0145] Wherein, T(p,q) is the spatial correlation of the local area image local relative to the reference image St after being offset by p in the horizontal direction and q in the vertical direction, local(a,b) is the pixel value of the local area image local at position (a,b), and St(ap,bq) is the pixel value of the reference image St at position (ap,bq);

[0146] When the spatial correlation reaches the maximum value, the registration is completed and the visible light image is obtained.

[0147] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An adaptive distance measurement method for dangerous sources in substations based on binocular vision, characterized in that: include: Step 1: Control the binocular camera to obtain visible light images in the substation; Step 2: Based on the visible light image, obtain the ranging environment; Step 3: When the ranging environment requires infrared thermal imaging, the infrared thermal imager is controlled to obtain infrared thermal image data of different objects in the substation; Step 4: Fuse and analyze the visible light image and infrared thermal image data, and determine the distance measurement result of the hazard source; The step 2: obtaining the ranging environment based on the visible light image, includes: Based on the visible light image, the lighting condition analysis is performed to obtain lighting analysis items; the analysis items include: image brightness, image contrast, whether the image lighting is uniform, and image exposure; Obtain standard ranging environment condition determination factors; According to the illumination analysis item, determine the current ranging environment condition determination factor; If the current ranging environment determination factor matches the standard ranging environment determination factor, then the ranging environment does not require infrared thermal imaging; If there is no match between the current ranging environment determination factor and the standard ranging environment determination factor, the ranging environment requires infrared thermal imaging; The step of obtaining the standard ranging environment condition determination factor comprises: Obtain a target ranging record set; wherein the target ranging record set includes multiple target ranging records; A control group of target ranging records is obtained; wherein the control group is a ranging record that integrates binocular camera images and infrared sensor data in the target ranging records; Determine the difference in distance measurement results based on the target distance measurement records and the control group; Determine the illumination conditions of the target ranging records whose ranging result difference is less than or equal to a preset ranging result difference threshold, and use them as extraction conditions; Based on the extraction conditions and factors, a template is constructed to construct the factors for determining the standard ranging environment conditions; The step of obtaining a target ranging record set includes: Obtaining the first pre-selected target ranging record; Parsing the first pre-selected target ranging record, and obtaining parameter similarity between binocular camera parameters of a binocular camera used in the first pre-selected target ranging record and binocular camera parameters of a binocular camera for shooting a visible light image; If the parameter similarity is greater than or equal to a preset parameter similarity threshold, the first pre-selected target ranging record is used as the second pre-selected target ranging record; Traversing each second pre-selected target ranging record in turn, and obtaining condition similarities between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to other second pre-selected target ranging records; If the conditional similarities are all less than or equal to the preset conditional similarity threshold, the second pre-selected target ranging record being traversed is used as the target ranging record; If there is a condition similarity greater than a preset condition similarity threshold, stop calculating the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records; Determine whether the calculation record with conditional similarity greater than a preset conditional similarity threshold is calculated for the first time, if so, take the second pre-selected target ranging record being traversed as the target ranging record and continue to traverse the second pre-selected target ranging record that has not been traversed, if not, directly continue to traverse the second pre-selected target ranging record that has not been traversed; After the traversal is completed, all target distance measurement records are integrated to obtain a target distance measurement record set.

2. The binocular vision-based adaptive distance measurement method for hazardous sources in a substation according to claim 1, characterized in that: Step 1: Control the binocular camera to obtain visible light images in the substation, including: Get the station map inside the substation; Plan the binocular camera image acquisition trajectory based on the station map; According to the binocular camera image acquisition trajectory, the binocular camera is controlled to obtain the visible light image in the substation.

3. The binocular vision-based adaptive distance measurement method for dangerous sources in a substation as claimed in claim 2, characterized in that: Also includes: When controlling the binocular camera to obtain the visible light image in the substation, multiple binocular camera signals are registered to obtain the visible light image; Among them, the multi-channel binocular camera signals are registered to obtain visible light images, including: During registration, the spatial correlation between the reference image and the local area image is obtained. The calculation formula of the spatial correlation is as follows: Wherein, T(p,q) is the spatial correlation of the local area image local relative to the reference image St after being offset by p in the horizontal direction and q in the vertical direction, local(a,b) is the pixel value of the local area image local at position (a,b), and St(ap,bq) is the pixel value of the reference image St at position (ap,bq); When the spatial correlation reaches the maximum value, the registration is completed and the visible light image is obtained.

4. The binocular vision-based adaptive distance measurement method for hazardous sources in a substation according to claim 1, characterized in that: Also includes: Before stopping calculating the condition similarity between the illumination condition corresponding to the second pre-selected target ranging record being traversed and the illumination condition corresponding to the remaining second pre-selected target ranging records, determining whether the shooting angles of the second pre-selected target ranging record being traversed and the second pre-selected target ranging record corresponding to the calculated condition similarity are consistent; If they are consistent, stop calculating the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records, and continue traversing the second pre-selected target ranging records that have not been traversed; If they are inconsistent, continue to calculate the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records; If the condition similarity between the lighting condition corresponding to the second pre-selected target ranging record being traversed and the lighting condition corresponding to the last second pre-selected target ranging record for condition similarity calculation is less than or equal to a preset condition similarity threshold, or the condition similarity is greater than the preset condition similarity threshold and the corresponding shooting angles are inconsistent, the second pre-selected target ranging record being traversed is used as the target ranging record.

5. The adaptive distance measurement system for dangerous sources in substations based on binocular vision is characterized by: include: Binocular camera control subsystem, used to control the binocular camera to obtain visible light images in the substation; A ranging environment acquisition subsystem is used to acquire ranging environment conditions based on visible light images; The infrared thermal imager control subsystem is used to control the infrared thermal imager to obtain infrared thermal image data of different objects in the substation when the ranging environment requires infrared thermal imaging; The hazard source ranging subsystem is used to fuse and analyze visible light images and infrared thermal imaging data and determine the ranging results of the hazard source; The ranging environment acquisition subsystem acquires the ranging environment based on the visible light image, including: Based on the visible light image, the lighting condition analysis is performed to obtain lighting analysis items; the analysis items include: image brightness, image contrast, whether the image lighting is uniform, and image exposure; Obtain standard ranging environment condition determination factors; According to the illumination analysis item, determine the current ranging environment condition determination factor; If the current ranging environment determination factor matches the standard ranging environment determination factor, then the ranging environment does not require infrared thermal imaging; If there is no match between the current ranging environment determination factor and the standard ranging environment determination factor, the ranging environment requires infrared thermal imaging; The step of obtaining the standard ranging environment condition determination factor comprises: Obtain a target ranging record set; wherein the target ranging record set includes multiple target ranging records; A control group of target ranging records is obtained; wherein the control group is a ranging record that integrates binocular camera images and infrared sensor data in the target ranging records; Determine the difference in distance measurement results based on the target distance measurement records and the control group; Determine the illumination conditions of the target ranging records whose ranging result difference is less than or equal to a preset ranging result difference threshold, and use them as extraction conditions; Based on the extraction conditions and factors, a template is constructed to construct the factors for determining the standard ranging environment conditions; The step of obtaining a target ranging record set includes: Obtaining the first pre-selected target ranging record; Parsing the first pre-selected target ranging record, and obtaining parameter similarity between binocular camera parameters of a binocular camera used in the first pre-selected target ranging record and binocular camera parameters of a binocular camera for shooting a visible light image; If the parameter similarity is greater than or equal to a preset parameter similarity threshold, the first pre-selected target ranging record is used as the second pre-selected target ranging record; Traversing each second pre-selected target ranging record in turn, and obtaining condition similarities between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to other second pre-selected target ranging records; If the conditional similarities are all less than or equal to the preset conditional similarity threshold, the second pre-selected target ranging record being traversed is used as the target ranging record; If there is a condition similarity greater than a preset condition similarity threshold, stop calculating the condition similarity between the illumination conditions corresponding to the second pre-selected target ranging record being traversed and the illumination conditions corresponding to the remaining second pre-selected target ranging records; Determine whether the calculation record with conditional similarity greater than a preset conditional similarity threshold is calculated for the first time, if so, take the second pre-selected target ranging record being traversed as the target ranging record and continue to traverse the second pre-selected target ranging record that has not been traversed, if not, directly continue to traverse the second pre-selected target ranging record that has not been traversed; After the traversal is completed, all target distance measurement records are integrated to obtain a target distance measurement record set.

6. The binocular vision-based adaptive distance measurement system for dangerous sources in substations as claimed in claim 5, characterized in that: Binocular camera control subsystem, including: The station map acquisition module is used to obtain the station map of the substation; The acquisition trajectory planning module is used to plan the binocular camera image acquisition trajectory according to the station map; The visible light image acquisition module is used to control the binocular camera to acquire the visible light image in the substation according to the binocular camera image acquisition trajectory.

7. The binocular vision-based adaptive distance measurement system for dangerous sources in substations according to claim 6, characterized in that: Also includes: The signal registration module is used to control the binocular camera to obtain the visible light image in the substation, and to register the signals of multiple binocular cameras to obtain the visible light image; Among them, the signal registration module performs the following operations: During registration, the spatial correlation between the reference image and the local area image is obtained. The calculation formula of the spatial correlation is as follows: Wherein, T(p,q) is the spatial correlation of the local area image local relative to the reference image St after being offset by p in the horizontal direction and q in the vertical direction, local(a,b) is the pixel value of the local area image local at position (a,b), and St(ap,bq) is the pixel value of the reference image St at position (ap,bq); When the spatial correlation reaches the maximum value, the registration is completed and the visible light image is obtained.

Images