Signal box safety detection system based on image analysis

Abstract

The invention discloses a signal box safety detection system based on image analysis, and the system comprises the steps: setting the collection time interval of each type of image, collecting the information of the corresponding type of image of a signal box after the collection time interval set by the corresponding type of image is reached, and carrying out the analysis, the cable evaluation index and the operation evaluation index corresponding to the signal box and the state evaluation index of each element are input into a pre-constructed database for comparison, and whether the connection cable, the element operation state and the element use state of the signal box are abnormal or not is judged according to the comparison result. The problem that in the prior art, detection and evaluation of operation states and use states of signal box connecting cables and signal box elements are lacked is solved, so that the safety state of the signal box is evaluated comprehensively and accurately, and the reliability and stability of the signal box are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of signal box safety detection, and in particular to a signal box safety detection system based on image analysis. Background Art

[0002] Signal boxes are key equipment in many industrial and communication systems, and their safe and stable operation is crucial to the normal operation of the entire system.

[0003] However, the signal box safety monitoring system in the prior art still has the following shortcomings: On the one hand, there is a lack of inspection and evaluation of the cables connecting the signal box. As an important part of the connection between the signal box and external equipment, the cables are exposed to complex environments for a long time and are prone to aging, damage, and bulges, causing equipment failure or downtime. On the other hand, there is a lack of detection and evaluation of the operating status and usage status of the signal box components. The internal components of the signal box will generate heat during operation. If the heat cannot be dissipated in time, the components will overheat and be damaged, affecting the overall performance of the signal box. The components themselves may also experience problems such as wear and deformation due to long-term use, which will gradually reduce the performance of the components and eventually cause failures.

[0004] In summary, a signal box safety detection system based on image analysis is introduced. Summary of the invention

[0005] In view of this, the present invention provides a signal box safety detection system based on image analysis to solve the problems raised by the above background technology.

[0006] The purpose of the present invention can be achieved through the following technical solutions: A signal box safety detection system based on image analysis, comprising: Image acquisition module: sets the acquisition time interval of each type of image, and after reaching the acquisition time interval set for the corresponding type of image, collects the corresponding type of image information of the signal box and sends it to the image analysis module; each type of image includes cable image, thermal radiation image and internal image; Image analysis module: receives various types of images of the signal box and performs preprocessing. After the preprocessing is completed, the various types of images of the signal box are comprehensively evaluated to obtain the cable evaluation index grea, the operation evaluation index greb and the status evaluation index greci of each component, and sends them to the abnormality processing module; where i represents the number corresponding to each component, i=1,2,......,t, t is the total number of components; Abnormal processing module: receives the cable evaluation index grea, operation evaluation index greb and status evaluation index greci of each component corresponding to each type of image of the signal box; inputs them into the pre-built database for comparison, and executes corresponding steps based on the comparison results.

[0007] In some embodiments, the cable image of the signal box is analyzed, specifically: Acquire the cable image of the signal box, and convert the cable image into the HSV color space. After the conversion is completed, extract the pre-stored cable aging reference image, and determine the aging color reference range of the cable aging reference image in the HSV color space; According to the determined aging color reference range, a corresponding threshold range is set for each channel in the HSV color space. After the setting is completed, for the acquired cable image, each pixel in the cable image is traversed to determine whether the value in the HSV color space is within the corresponding set threshold range, the number of pixels within the threshold range is counted, and the actual area is converted based on the resolution of the image. After the conversion is completed, the aging area value f1 corresponding to the cable image is obtained; Perform edge detection on the cable image, and identify the damaged position in the edge image corresponding to the cable image after detection; count pixels along the damaged position, and use the counted number of pixels as the number of damaged pixels corresponding to the current damaged position. After converting the actual area based on the resolution of the image, the fracture area of ​​the current damaged position is obtained, and the fracture areas of each group of identified damaged positions are accumulated to obtain the damaged area value f2 corresponding to the cable image; Extract the edge curve of the edge image corresponding to the cable image, and determine the raised area for the edge curve of the edge image corresponding to the cable image; select the vertical pixel distance between the highest point of each raised area and the adjacent normal edge point, and convert the actual height according to the resolution of the image to obtain the raised height value corresponding to each raised area; The allowable value corresponding to the preset bulge height value; compare the bulge height value corresponding to each bulge area with the allowable value, if the bulge height value corresponding to a group of bulge areas is greater than the preset allowable value, then mark the bulge area as an abnormal area; for each group of abnormal areas, determine the edge contour of each group of abnormal areas, and count the number of pixels in the corresponding contour of each abnormal area, and perform actual area conversion based on the resolution of the image and then sum them up to obtain the bulge area value f3 corresponding to the cable image.

[0008] In some embodiments, the cable evaluation index grea is obtained, specifically: According to the formula The aging area value f1, the damaged area value f2 and the raised area value f3 of the cable image are weighted to obtain the cable evaluation index grea of ​​the signal box; , as well as They respectively represent the allowable values ​​of aging area, damaged area and convex area; a1, a2 and a3 are the influencing weight factors of the aging area value f1, damaged area value f2 and convex area value f3.

[0009] In some embodiments, the thermal radiation image of the signal box is analyzed to obtain the operation evaluation index greb, specifically: Acquire a thermal radiation image of the signal box, and divide the thermal radiation image into regions according to preset functional division regions to obtain each divided functional region corresponding to the thermal radiation image of the signal box; Set a weight coefficient corresponding to each divided functional area. For each divided functional area, traverse the pixels in each divided functional area and extract the temperature value, calculate the average temperature of the pixels in each divided functional area, and use the calculated average values ​​of each group as the temperature performance value of each divided functional area. Preset the maximum allowable value corresponding to the temperature performance value of each divided functional area; calculate the ratio of the temperature performance value of each divided functional area to the corresponding preset maximum allowable value, multiply the calculated ratio by the corresponding set weight coefficient, and then sum them up to obtain the operation evaluation index greb of the signal box; that is, the temperature performance value of each divided functional area / the corresponding maximum allowable value.

[0010] In some embodiments, the internal image of the signal box is analyzed to obtain the state evaluation index greci of each component, specifically: Acquire the internal image of the signal box, and extract the component area image corresponding to each component from the internal image, for each component area image describing the outline of the component, extract the standard outline of each component from a pre-built database, calculate the difference between the outline area described by each component and the extracted standard outline area, and use the standard outline area of ​​each component to subtract the corresponding described outline area to obtain the outline difference value Eqi corresponding to each component; Convert each component region image into a grayscale image, and extract the wear region from each component region image using a preset threshold; The pixel count of the extracted wear area is performed, and after the statistics are completed, the actual area is converted according to the resolution of the image to obtain the wear quantification value Eri corresponding to each component; According to the formula The contour difference value Eqi and the wear quantification value Eri corresponding to each component are weightedly calculated to obtain the state evaluation index greci of each component; and They respectively represent the allowable value of contour difference and the allowable value of wear quantification corresponding to each component. b1 and b2 are the influencing weight factors of the contour difference value Eqi and the wear quantification value Er, respectively.

[0011] In some embodiments, corresponding steps are performed based on the comparison results, specifically: S1: extract the reference threshold index corresponding to the cable evaluation index grea from the pre-built database, and compare the cable evaluation index grea corresponding to the cable image of the signal box with the reference threshold index. If it is greater than the reference threshold index, the cable abnormality signaling is triggered and a circle is drawn with the location of the signal box as the center and the distance as the radius. The maintenance personnel closest to the signal box within the circle at the current trigger time point are obtained as the processing personnel, and the line abnormality signaling is sent to the mobile terminal of the processing personnel; S2: extract the reference threshold index corresponding to the operation evaluation index greb from the pre-built database, and compare the operation evaluation index greb corresponding to the thermal radiation image of the signal box with the reference threshold index. If it is greater than the reference threshold index, the temperature abnormality signal is triggered and the operation of the signal box is stopped. Similarly, after selecting the processing personnel in step S1, the temperature abnormality signal is sent to the mobile terminal of the processing personnel; S3: Extract the reference threshold index corresponding to each component status assessment index greci from the pre-constructed database, and compare the status assessment index greci of each component corresponding to the internal image of the signal box with the corresponding reference threshold index. If the status assessment index greci of a group of components is greater than the corresponding reference threshold index, the component hidden danger signaling is triggered with the component number as the header. After the triggering, the processing personnel are selected in the same way as step S1, and the component hidden danger signaling with the component number as the header is sent to the processing personnel's mobile terminal.

[0012] In some embodiments, if the running evaluation index greb is less than the reference threshold index in step S2, the following is specifically performed: S2-201: If the operation evaluation index greb is less than the reference threshold index, taking the current time point as the starting point, continuously collecting the thermal radiation images of the signal box at each time point within the set time period after the starting point, and analyzing to obtain the operation evaluation index gred of the signal box at each time point within the set time period after the starting point; The operation evaluation index gred at the current time point and the operation evaluation index gred at each time point are arranged in chronological order. After the arrangement is completed, the difference between two adjacent groups of operation evaluation index gred is calculated. The difference calculation order is the operation evaluation index on the right of the two adjacent groups minus the operation evaluation index on the left; If a group of differences is a negative value, then the absolute value of the group of differences is taken and marked as the adjacent descending value; if a group of differences is a positive value, then the group of differences is marked as the adjacent ascending value; sum up all the adjacent descending values ​​and adjacent ascending values ​​respectively to obtain the adjacent ascending total value and the adjacent descending total value; The change ratio is obtained by calculating the ratio between the total value of adjacent increases and the total value of adjacent decreases, that is, by calculating the total value of adjacent increases / the total value of adjacent decreases; The maximum allowable ratio corresponding to the preset change ratio is compared with the preset maximum allowable value within the set time period. If the change ratio is greater than the corresponding maximum allowable ratio, an abnormal trend signaling is triggered; an additional coefficient is set corresponding to the abnormal trend signaling; The operation evaluation index gred at the end of the arrangement is extracted and multiplied by the obtained additional coefficient to obtain the hidden danger assessment index, and the hidden danger assessment index is compared with the reference threshold index corresponding to the operation evaluation index greb. If it is greater than the corresponding reference threshold index, the temperature abnormality signaling is also triggered.

[0013] Compared with the prior art, the present invention has the following beneficial effects: The present invention sets the collection time interval of each type of image, and after reaching the collection time interval set for the corresponding type of image, collects and analyzes the corresponding type of image information of the signal box, thereby obtaining the cable evaluation index, operation evaluation index and status evaluation index of each component corresponding to the signal box, and inputs them into a pre-built database for comparison. According to the comparison result, it is determined whether the connection cables, component operation status and component use status of the signal box are abnormal, which solves the problem of the lack of detection and evaluation of the connection cables of the signal box, the operation status of the signal box components and the use status in the prior art, thereby comprehensively and accurately evaluating the safety status of the signal box and improving the reliability and stability of the signal box. The present invention inputs the cable evaluation index, operation evaluation index and status evaluation index of each component corresponding to the signal box into a pre-built database, triggers the corresponding signaling and processes it based on the comparison result, so as to avoid further deterioration of the fault and affect the normal operation of the signal box; When the operation evaluation index is less than the reference threshold index, the present invention extracts the operation evaluation index of the signal box at each time point within the set time period after the starting point, and analyzes it to obtain the change ratio. If the change ratio is greater than the corresponding maximum allowable ratio, the abnormal trend signal is triggered, the operation evaluation index arranged at the end is extracted and multiplied with the obtained additional coefficient to obtain the hidden danger assessment index, and the hidden danger assessment index is compared with the reference threshold index corresponding to the operation evaluation index. If it is greater than the corresponding reference threshold index, the temperature abnormality signal is also triggered, which helps to discover potential hidden dangers that have not yet caused obvious failures but may develop into serious problems in the future in advance, and provides a strong basis for preventive maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Further details, features and advantages of the present application are disclosed in the following description of exemplary embodiments in conjunction with the accompanying drawings, in which: Figure 1 It is a principle block diagram of the present invention. DETAILED DESCRIPTION

[0015] Several embodiments of the present application will be described in more detail below with reference to the accompanying drawings so that those skilled in the art can implement the present application. The present application can be embodied in many different forms and purposes and should not be limited to the embodiments described herein. These embodiments are provided to make the present application comprehensive and complete, and to fully convey the scope of the present application to those skilled in the art. The embodiments do not limit the present application.

[0016] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those of ordinary skill in the art to which this application belongs. It will be further understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the relevant art and / or the context of this specification, and will not be interpreted in an idealized or overly formal sense unless explicitly defined as such herein.

[0017] See also Figure 1 As shown, a signal box safety detection system based on image analysis includes an image acquisition module, an image analysis module and an abnormality processing module; The image acquisition module is used to set the acquisition time interval of each type of image, and after reaching the acquisition time interval set for the corresponding type of image, collect the corresponding type of image information of the signal box and send it to the image analysis module; the various types of images include cable images, thermal radiation images and internal images; Set up regular high-definition cameras, thermal imaging cameras, and endoscope cameras at locations around the signal box to ensure that they can fully cover the entirety of the signal box's connecting cables and key internal locations; Set the time interval for image acquisition or use real-time acquisition mode to obtain image data samples; For ordinary high-definition cameras, according to the set collection time interval, for example, once every 3 days, the connection cable of the signal box is collected. The thermal imaging camera collects the thermal radiation image of the signal box according to the set collection time interval, for example, once every 30 minutes, to reflect its temperature distribution; The endoscopic camera collects detailed images of components inside the signal box according to a set collection time interval, for example, once a day; The image analysis module is used to receive various types of images from the signal box and perform preprocessing; the preprocessing includes image denoising and enhancement, etc. After the preprocessing is completed, the various types of images of the signal box are comprehensively evaluated to obtain the cable evaluation index grea, the operation evaluation index greb and the status evaluation index greci of each component, and send them to the abnormality processing module; where i represents the number corresponding to each component, i=1,2,......,t, t is the total number of components; The cable evaluation index grea is obtained by comprehensively evaluating the signal box cable image; Comprehensively evaluate the thermal radiation image of the signal box to obtain the operation evaluation index greb; Comprehensively evaluate the internal image of the signal box to obtain the condition evaluation index greci of each component; The cable image of the signal box is analyzed to obtain the cable evaluation index grea, which is: Acquire the cable image of the signal box, and convert the cable image into the HSV color space. After the conversion is completed, extract the pre-stored cable aging reference image, and determine the aging color reference range of the cable aging reference image in the HSV color space; According to the determined aging color reference range, a corresponding threshold range is set for each channel in the HSV color space. After the setting is completed, for the acquired cable image, each pixel in the cable image is traversed to determine whether the value in the HSV color space is within the corresponding set threshold range, the number of pixels within the threshold range is counted, and the actual area is converted based on the resolution of the image. After the conversion is completed, the aging area value f1 corresponding to the cable image is obtained; Use the Canny detection algorithm to perform edge detection on the cable image. After detection, identify the damaged position in the edge image corresponding to the cable image; that is, the place where the edge is broken and discontinuous; and count the pixels along the broken position, and use the counted number of pixels as the number of damaged pixels corresponding to the current damaged position. After converting the actual area based on the resolution of the image, the broken area of ​​the current damaged position is obtained, and the broken areas of each group of identified damaged positions are accumulated to obtain the damaged area value f2 corresponding to the cable image; Extract the edge curve of the edge image corresponding to the cable image, and determine the convex area for the edge curve of the edge image corresponding to the cable image; find the convex part by observing the change of the edge curve; select the vertical pixel distance between the highest point of each convex area and the adjacent normal edge point, and convert the actual height according to the resolution of the image to obtain the convex height value corresponding to each convex area; According to the normal cable edge curve pre-constructed in the database, the allowable value corresponding to the convex height value is preset; the convex height value corresponding to each convex area is compared with the allowable value, and if the convex height value corresponding to a group of convex areas is greater than the preset allowable value, the convex area is marked as an abnormal area; for each group of abnormal areas, the edge contour of each group of abnormal areas is determined, and the number of pixels in the corresponding contour of each abnormal area is counted, and the actual area conversion is performed based on the resolution of the image and then summed to obtain the convex area value f3 corresponding to the cable image; According to the formula The aging area value f1, the damaged area value f2 and the raised area value f3 of the cable image are weighted to obtain the cable evaluation index grea of ​​the signal box; , as well as They represent the allowable values ​​of aging area, damaged area and raised area respectively; they are preset by technical personnel and can be dynamically adjusted according to actual conditions; a1, a2 and a3 are the influence weight factors of aging area value f1, damaged area value f2 and raised area value f3 respectively; It should be noted that by analyzing and quantifying various common and critical status characteristics of cables, such as cable aging, breakage, and bulges, the actual condition of the cables can be fully presented in a digital way.

[0018] The thermal radiation image of the signal box is analyzed to obtain the operation evaluation index greb, which is: Acquire the thermal radiation image of the signal box, and divide the thermal radiation image into regions according to the preset functional division areas to obtain the functional division areas corresponding to the thermal radiation image of the signal box; divide the areas according to the functional categories of the components, for example, divide all components related to signal transmission (such as signal lines, interfaces, amplifiers, etc.) into a "signal transmission area"; and classify components involved in power supply and conversion (such as transformers, rectifiers, power chips, etc.) into a "power supply area". This division helps to focus on the temperature characteristics of different functional components and their impact on the overall function when analyzing the thermal radiation image; Set a weight coefficient corresponding to each divided functional area, and for each divided functional area, traverse the pixels in each divided functional area and extract the temperature value; use image analysis software or programming language to implement this operation; calculate the average temperature of the pixels in each divided functional area, and use the calculated average values ​​of each group as the temperature representation value of each divided functional area; Preset the maximum allowable value corresponding to the temperature performance value of each divided functional area; calculate the ratio of the temperature performance value of each divided functional area to the corresponding preset maximum allowable value, and multiply the calculated ratio by the corresponding set weight coefficient, and then sum them up to obtain the operation evaluation index greb of the signal box; that is, the temperature performance value of each divided functional area / the corresponding maximum allowable value; It should be noted that by dividing and analyzing the thermal radiation image of the signal box according to functional areas, the thermal conditions inside the signal box can be grasped as a whole. It is no longer a simple observation of local temperature, but a comprehensive consideration of the thermal performance of each functional module.

[0019] The internal image of the signal box is analyzed to obtain the condition evaluation index greci of each component, which is: Obtain an internal image of the signal box, and extract the component area image corresponding to each component from the internal image. For each component area image, describe the contour of the component by an edge detection algorithm, extract the standard contour of each component from a pre-built database, calculate the difference between the contour area described by each component and the extracted standard contour area, and subtract the corresponding contour area described from the standard contour area of ​​each component to obtain the contour difference value Eqi corresponding to each component; Convert each component area image into a grayscale image, and extract the wear area from each component area image through a preset threshold; in the wear area, due to the uneven surface, the standard deviation of the grayscale difference will be larger than that of the normal area. By setting a suitable threshold, the possible wear area can be preliminarily screened out; The pixel count of the extracted wear area is performed, and after the statistics are completed, the actual area is converted according to the resolution of the image to obtain the wear quantification value Eri corresponding to each component; According to the formula The contour difference value Eqi and the wear quantification value Eri corresponding to each component are weightedly calculated to obtain the state evaluation index greci of each component; and They represent the contour difference allowable value and wear quantification allowable value corresponding to each element, b1 and b2 are the influence weight factors of the contour difference value Eqi and the wear quantification value Er, respectively. The specific values ​​are set by the technicians, and b2>b1; It should be noted that this method comprehensively considers two important dimensions, namely the change in component contour and wear condition, to evaluate the component status. It no longer relies solely on a single qualitative observation or a certain indicator, but quantifies the physical morphological changes and surface wear degree of the component by calculating the contour difference value and the wear quantification value, which can more comprehensively and accurately reflect the actual condition of the component.

[0020] The exception handling module is used to receive the cable evaluation index grea, the operation evaluation index greb and the status evaluation index greci of each component corresponding to each type of image of the signal box; and input them into the pre-built database for comparison, and execute corresponding steps based on the comparison results; Specifically: S1: extract the reference threshold index corresponding to the cable evaluation index grea from the pre-built database, and compare the cable evaluation index grea corresponding to the cable image of the signal box with the reference threshold index. If it is greater than the reference threshold index, the cable abnormality signaling is triggered and a circle is drawn with the location of the signal box as the center and the distance as the radius. The maintenance personnel closest to the signal box within the circle at the current trigger time point are obtained as the processing personnel, and the line abnormality signaling is sent to the mobile terminal of the processing personnel; S2: extract the reference threshold index corresponding to the operation evaluation index greb from the pre-built database, and compare the operation evaluation index greb corresponding to the thermal radiation image of the signal box with the reference threshold index. If it is greater than the reference threshold index, the temperature abnormality signal is triggered and the operation of the signal box is stopped. Similarly, after selecting the processing personnel in step S1, the temperature abnormality signal is sent to the mobile terminal of the processing personnel; S2-201: If the operation evaluation index greb is less than the reference threshold index, taking the current time point as the starting point, continuously collecting the thermal radiation images of the signal box at each time point within the set time period after the starting point, and analyzing to obtain the operation evaluation index gred of the signal box at each time point within the set time period after the starting point; The operation evaluation index gred at the current time point and the operation evaluation index gred at each time point are arranged in chronological order. After the arrangement is completed, the difference between two adjacent groups of operation evaluation index gred is calculated. The difference calculation order is the operation evaluation index on the right of the two adjacent groups minus the operation evaluation index on the left; If a group of differences is a negative value, then the absolute value of the group of differences is taken and marked as the adjacent descending value; if a group of differences is a positive value, then the group of differences is marked as the adjacent ascending value; sum up all the adjacent descending values ​​and adjacent ascending values ​​respectively to obtain the adjacent ascending total value and the adjacent descending total value; The change ratio is obtained by calculating the ratio between the total value of adjacent increases and the total value of adjacent decreases, that is, by calculating the total value of adjacent increases / the total value of adjacent decreases; It should be noted that if the total value of adjacent decreases or the total value of adjacent increases is 0, the integer 1 is taken; The maximum allowable ratio corresponding to the preset change ratio is compared with the preset maximum allowable value within the set time period. If the change ratio is greater than the corresponding maximum allowable ratio, the abnormal trend signaling is triggered; an additional coefficient corresponding to the abnormal trend signaling is set; the additional coefficient is set to 1.175 and is set by the technical staff; Extract the operation evaluation index gred at the end of the arrangement and multiply it with the obtained additional coefficient to obtain the hidden danger evaluation index, and compare the hidden danger evaluation index with the reference threshold index corresponding to the operation evaluation index greb. If it is greater than the corresponding reference threshold index, the temperature abnormality signaling is also triggered; S3: extracting the reference threshold indexes corresponding to the component status evaluation indexes greci from the pre-built database, and comparing the status evaluation indexes greci of the components corresponding to the internal image of the signal box with the corresponding reference threshold indexes. If the status evaluation index greci of a group of components is greater than the corresponding reference threshold index, the component hidden danger signaling is triggered with the component number as the header. After the triggering, after selecting the processing personnel in step S1, the component hidden danger signaling with the component number as the header is sent to the processing personnel's mobile terminal; It should be noted that by comparing the evaluation index corresponding to each type of image of the signal box with the reference threshold index in the database, it is possible to quickly determine whether an abnormal situation occurs; once the index exceeds the reference threshold index, the corresponding abnormal signaling is immediately triggered, thereby achieving timely detection and early warning of the fault, which helps to avoid further deterioration of the fault and affect the normal operation of the signal box; For the operation evaluation index greb, when it is less than the reference threshold index, it is not simply judged as no abnormality, but further continuously collects thermal radiation images at each time point within the set time period and analyzes them to obtain the corresponding operation evaluation index gred; by arranging the indexes at different time points, calculating adjacent differences, summing them, and calculating the change ratio, the operation status change trend of the signal box during this period can be deeply analyzed; By presetting the maximum allowable ratio of the change ratio, when the calculated change ratio exceeds the allowable ratio, the abnormal trend signal is triggered, indicating that there may be a potential abnormal operation trend. And further by calculating the hidden danger assessment index and comparing it with the reference threshold index, even if the current operation assessment index greb does not exceed the threshold, if the hidden danger assessment index exceeds the reference threshold index, the temperature abnormality signal is also triggered, which helps to discover potential hidden dangers that have not yet caused obvious failures but may develop into serious problems in the future in advance, providing a strong basis for preventive maintenance; The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A signal box safety detection system based on image analysis, characterized in that: include: Image acquisition module: sets the acquisition time interval of each type of image, and after reaching the acquisition time interval set for the corresponding type of image, collects the corresponding type of image information of the signal box and sends it to the image analysis module; each type of image includes cable image, thermal radiation image and internal image; Image analysis module: receives various types of images from the signal box and performs preprocessing. After the preprocessing is completed, a comprehensive evaluation is performed on the various types of images of the signal box to obtain the cable evaluation index grea, the operation evaluation index greb and the status evaluation index greci of each component, and sends them to the exception handling module; Where i represents the number corresponding to each component, i=1,2,......,t, t is the total number of components; Abnormal processing module: receives the cable evaluation index grea, operation evaluation index greb and status evaluation index greci of each component corresponding to each type of image of the signal box; inputs them into the pre-built database for comparison, and executes corresponding steps based on the comparison results.

2. The signal box safety detection system based on image analysis according to claim 1 is characterized in that: Analyze the cable image of the signal box, specifically: Acquire the cable image of the signal box, and convert the cable image into the HSV color space. After the conversion is completed, extract the pre-stored cable aging reference image, and determine the aging color reference range of the cable aging reference image in the HSV color space; According to the determined aging color reference range, a corresponding threshold range is set for each channel in the HSV color space. After the setting is completed, for the acquired cable image, each pixel in the cable image is traversed to determine whether the value in the HSV color space is within the corresponding set threshold range, the number of pixels within the threshold range is counted, and the actual area is converted based on the resolution of the image. After the conversion is completed, the aging area value f1 corresponding to the cable image is obtained; Perform edge detection on the cable image, and identify the damaged position in the edge image corresponding to the cable image after detection; count pixels along the damaged position, and use the counted number of pixels as the number of damaged pixels corresponding to the current damaged position. After converting the actual area based on the resolution of the image, the fracture area of ​​the current damaged position is obtained, and the fracture areas of each group of identified damaged positions are accumulated to obtain the damaged area value f2 corresponding to the cable image; Extract the edge curve of the edge image corresponding to the cable image, and determine the raised area for the edge curve of the edge image corresponding to the cable image; select the vertical pixel distance between the highest point of each raised area and the adjacent normal edge point, and convert the actual height according to the resolution of the image to obtain the raised height value corresponding to each raised area; The allowable value corresponding to the preset bulge height value; compare the bulge height value corresponding to each bulge area with the allowable value, if the bulge height value corresponding to a group of bulge areas is greater than the preset allowable value, then mark the bulge area as an abnormal area; for each group of abnormal areas, determine the edge contour of each group of abnormal areas, and count the number of pixels in the corresponding contour of each abnormal area, and perform actual area conversion based on the resolution of the image and then sum them up to obtain the bulge area value f3 corresponding to the cable image.

3. The signal box safety detection system based on image analysis according to claim 2 is characterized in that: Get the cable evaluation index grea, specifically: According to the formula The aging area value f1, the damaged area value f2 and the raised area value f3 of the cable image are weighted to obtain the cable evaluation index grea of ​​the signal box; , as well as They respectively represent the allowable values ​​of aging area, damaged area and convex area; a1, a2 and a3 are the influencing weight factors of the aging area value f1, damaged area value f2 and convex area value f3.

4. The signal box safety detection system based on image analysis according to claim 3 is characterized in that: The thermal radiation image of the signal box is analyzed to obtain the operation evaluation index greb, which is: Acquire a thermal radiation image of the signal box, and divide the thermal radiation image into regions according to preset functional division regions to obtain each divided functional region corresponding to the thermal radiation image of the signal box; Set a weight coefficient corresponding to each divided functional area. For each divided functional area, traverse the pixels in each divided functional area and extract the temperature value, calculate the average temperature of the pixels in each divided functional area, and use the calculated average values ​​of each group as the temperature performance value of each divided functional area. Preset the maximum allowable value corresponding to the temperature performance value of each divided functional area; calculate the ratio of the temperature performance value of each divided functional area to the corresponding preset maximum allowable value, multiply the calculated ratio by the corresponding set weight coefficient, and then sum them up to obtain the operation evaluation index greb of the signal box; that is, the temperature performance value of each divided functional area / the corresponding maximum allowable value.

5. The signal box safety detection system based on image analysis according to claim 4 is characterized in that: The internal image of the signal box is analyzed to obtain the condition evaluation index greci of each component, which is: Acquire the internal image of the signal box, and extract the component area image corresponding to each component from the internal image, for each component area image describing the outline of the component, extract the standard outline of each component from a pre-built database, calculate the difference between the outline area described by each component and the extracted standard outline area, and use the standard outline area of ​​each component to subtract the corresponding described outline area to obtain the outline difference value Eqi corresponding to each component; Convert each component region image into a grayscale image, and extract the wear region from each component region image using a preset threshold; The pixel count of the extracted wear area is performed, and after the statistics are completed, the actual area is converted according to the resolution of the image to obtain the wear quantification value Eri corresponding to each component; According to the formula The contour difference value Eqi and the wear quantification value Eri corresponding to each component are weightedly calculated to obtain the state evaluation index greci of each component; and They respectively represent the allowable value of contour difference and the allowable value of wear quantification corresponding to each component. b1 and b2 are the influencing weight factors of the contour difference value Eqi and the wear quantification value Er, respectively.

6. The signal box safety detection system based on image analysis according to claim 5 is characterized in that: Based on the comparison results, the corresponding steps are performed, specifically: S1: extract the reference threshold index corresponding to the cable evaluation index grea from the pre-built database, and compare the cable evaluation index grea corresponding to the cable image of the signal box with the reference threshold index. If it is greater than the reference threshold index, the cable abnormality signaling is triggered and a circle is drawn with the location of the signal box as the center and the distance as the radius. The maintenance personnel closest to the signal box within the circle at the current trigger time point are obtained as the processing personnel, and the line abnormality signaling is sent to the mobile terminal of the processing personnel; S2: extract the reference threshold index corresponding to the operation evaluation index greb from the pre-built database, and compare the operation evaluation index greb corresponding to the thermal radiation image of the signal box with the reference threshold index. If it is greater than the reference threshold index, the temperature abnormality signal is triggered and the operation of the signal box is stopped. Similarly, after selecting the processing personnel in step S1, the temperature abnormality signal is sent to the mobile terminal of the processing personnel; S3: Extract the reference threshold index corresponding to each component status assessment index greci from the pre-constructed database, and compare the status assessment index greci of each component corresponding to the internal image of the signal box with the corresponding reference threshold index. If the status assessment index greci of a group of components is greater than the corresponding reference threshold index, the component hidden danger signaling is triggered with the component number as the header. After the triggering, the processing personnel are selected in the same way as step S1, and the component hidden danger signaling with the component number as the header is sent to the processing personnel's mobile terminal.

7. The signal box safety detection system based on image analysis according to claim 6 is characterized in that: In step S2, if the running evaluation index greb is less than the reference threshold index, the following specific steps are performed: S2-201: If the operation evaluation index greb is less than the reference threshold index, taking the current time point as the starting point, continuously collecting the thermal radiation images of the signal box at each time point within the set time period after the starting point, and analyzing to obtain the operation evaluation index gred of the signal box at each time point within the set time period after the starting point; The operation evaluation index gred at the current time point and the operation evaluation index gred at each time point are arranged in chronological order. After the arrangement is completed, the difference between two adjacent groups of operation evaluation index gred is calculated. The difference calculation order is the operation evaluation index on the right of the two adjacent groups minus the operation evaluation index on the left; If a group of differences is a negative value, the absolute value of the group of differences is taken and marked as the adjacent descending value; if a group of differences is a positive value, the group of differences is marked as the adjacent ascending value; Sum all adjacent decreasing values ​​and adjacent increasing values ​​respectively to obtain the total value of adjacent increasing values ​​and the total value of adjacent decreasing values; The change ratio is obtained by calculating the ratio between the total value of adjacent increases and the total value of adjacent decreases, that is, by calculating the total value of adjacent increases / the total value of adjacent decreases; The maximum allowable ratio corresponding to the preset change ratio is compared with the preset maximum allowable value within the set time period. If the change ratio is greater than the corresponding maximum allowable ratio, an abnormal trend signaling is triggered; Set an additional coefficient corresponding to the abnormal trend signal; The operation evaluation index gred at the end of the arrangement is extracted and multiplied by the obtained additional coefficient to obtain the hidden danger assessment index, and the hidden danger assessment index is compared with the reference threshold index corresponding to the operation evaluation index greb. If it is greater than the corresponding reference threshold index, the temperature abnormality signaling is also triggered.

Images