A method for detecting thermal information of brake disc surfaces under dynamic operating conditions using computer processing.

JP2025501759A5Pending Publication Date: 2025-12-25FRENI BREMBO S P A O PIU BREVEMENTE BREMBO
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
JP2024538141
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Priority Date
2021-12-23
Filing Date
2022-12-19
Publication Date
2025-12-25

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

A method for detecting thermal information of a brake disc surface under dynamic operating conditions by computer processing. The method for automatically detecting information about temperature changes on a brake disc surface under dynamic operating conditions and / or about the location / distribution of hot spots and / or bands and / or zones by computer processing comprises the steps of acquiring a plurality of digital images and / or digital videos, acquiring a representation of each frame of the digital images and / or digital videos in the form of matrix data, and performing an algorithmic process consisting of an image analysis or "computer vision" algorithm on the basis of said matrix data to acquire and provide information about temperature changes on an area of ​​interest on the surface of the brake disc and / or about the location / distribution of hot spots and / or bands and / or zones. The acquiring step comprises acquiring by at least one thermal imaging camera a plurality of digital images and / or a digital video consisting of a plurality of digital video frames of at least one area of ​​interest of the brake disc to be thermally characterized during said change of dynamic operating conditions. Each such digital image and / or digital video frame depicts the temperature detected at each point by means of a color map. The processing step includes identifying a region of interest by generating a mask based on an analysis of the plurality of acquired digital images and / or digital video frames, then applying the generated mask to each of the acquired digital images and / or digital video frames to obtain processed matrix data related to the region of interest, then defining an analysis region and performing a thermal analysis on the data related to the defined analysis region to obtain information regarding temperature changes at each point of interest, and finally identifying hot points based on the information regarding temperature changes at each point of interest. The providing step includes providing the obtained information in the form of a digital image or graphic and / or a digital table and / or a digital video.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a method for automatically detecting thermal information of a brake disc surface under dynamic operating conditions by computer processing.

[0002] In particular, the invention relates to a method for automatically detecting information regarding temperature variations on the surface of a brake disc for a vehicle braking system and / or information regarding the location / distribution of hot spots and / or bands and / or zones during dynamic bench testing. [Background technology]

[0003] The dissipation of heat generated during braking is one of the most important issues in evaluating the performance of a braking system.

[0004] In the case of brake discs, heat generated in the friction components (pads and braking surface) can cause thermal strains in the disc, leading to localized contact areas between mechanical elements and the development of thermal hot spots. Hot spots are areas of the braking band characterized by high thermal gradients.

[0005] The formation of thermal hot spots during braking can have several adverse effects.

[0006] From a materials perspective, it has been demonstrated that thermomechanical stresses due to the presence of thermal hot spots induce traction and compression stress cycles accompanied by plastic deformation fluctuations.

[0007] The presence of thermal hot spots also affects the appearance and propagation of cracks on the disk surface.

[0008] Finally, from a driving perspective, high temperatures within the braking system are known to deteriorate braking performance and induce the phenomenon of "fading" or "hot judder", i.e., undesirable low frequency vibrations.

[0009] It is extremely important to analyze the thermal hot spot phenomenon in brake discs under dynamic conditions, both to properly characterize the brake discs and to obtain suggestions for improving the design of the brake disc itself.

[0010] Currently, the standard procedure for studying the hot spot phenomenon is to test the braking system on a dynamometer test bench, where an experimental setup is provided so that a test device equipped with, for example, a thermograph can capture images representative of the temperature information.

[0011] In each test, a sequence of braking maneuvers predefined in terms of motion parameters is applied, and the frames acquired at each braking maneuver can be combined in sequence to form a video.

[0012] However, to analyze the relationship between the temperature distribution on the disk and complex physical phenomena, such as the occurrence of deformations or vibrations, the standard analysis tools provided by currently known test tools are at least partially missing or inadequate.

[0013] Indeed, although the standard software for known infrared thermography currently available makes it possible to automatically extract some information, such information is not sufficient to perform the above analysis in a satisfactory manner.

[0014] For this reason, the prior art in this field prescribes an individual visual inspection by a specialized operator of the video provided by the infrared thermal camera in order to record information about non-uniformities in the temperature distribution, such as hot spots, cold zones, hot bands, i.e. circular crowns of higher temperature compared to the surroundings, etc. From these features, metrics related to their number, size and temperature values ​​are recorded.

[0015] This known information extraction procedure is very time consuming since it requires a qualified operator to individually screen the captured video.

[0016] Furthermore, such procedures are usually subject to operator bias, resulting in less objective and less reproducible results.

[0017] As disclosed above, there remain many unmet needs in the area of ​​detection and analysis of information related to thermal hot spots on brake discs under dynamic operating conditions, and currently known solutions do not provide a sufficiently effective solution. Summary of the Invention

[0018] Accordingly, the present invention is directed to a method for detecting and analyzing thermal information associated with a brake disc under dynamic operating conditions.

[0019] In particular, it is an object of the present invention to provide a method for automatically detecting, by computer processing, information regarding temperature changes on the surface of a brake disc under dynamic operating conditions and / or information regarding the localization / distribution of hot spots and / or bands and / or zones, which makes it possible to at least partially avoid the drawbacks complained about above with reference to the prior art and to meet the aforementioned needs which are particularly felt in the technical field considered. Such an object is achieved by a method according to claim 1.

[0020] Further embodiments of such a method are defined in claims 2-18. [Brief description of the drawings]

[0021] Further features and advantages of the method according to the invention will become apparent from the following description of preferred embodiments, given by way of non-limiting indication, with reference to the attached drawings, in which:

[0022] [Figure 1]FIG. 1 shows an example of a framing obtained in a digital image / video acquisition step, provided in an embodiment of the method of the present invention.

[0023] [Diagram 2] FIG. 2 shows a simplified flow diagram of the steps involved in an embodiment of the method according to the invention.

[0024] [Diagram 3] FIG. 3 shows the hierarchy of analyses performed in an embodiment of the method according to the invention.

[0025] [Figure 4] FIG. 4 illustrates the different image processing steps provided in a mask definition procedure constituted by an embodiment of the method according to the invention.

[0026] [Diagram 5] FIG. 5 is a simplified flow diagram of the steps included in the "Data Preparation" block of FIG. 2, according to an embodiment.

[0027] [Figure 6] FIG. 6 shows an example of setting input parameters for hotspot search and validation according to an embodiment of the method according to the invention.

[0028] [Figure 7] FIG. 7 shows a simplified flow diagram of the steps included in the "Temperature Analysis" block of FIG. 2, according to implementation options.

[0029] [Figure 8] 8 and 9 respectively show a re-elaborate video and a line graph with connected frames provided as output in respective embodiments of the method according to the invention. [Figure 9] 8 and 9 respectively show a re-elaborate video and a line graph with connected frames provided as output in respective embodiments of the method according to the invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0030] With reference to Figures 1-9, a method for automatically detecting information regarding temperature changes and / or the location / distribution of hot spots and / or bands and / or zones on the surface of a brake disc under dynamic operating conditions by computer processing is described.

[0031] The method comprises the steps of acquiring a plurality of digital images and / or a digital video, acquiring a representation of each frame of the digital images and / or the digital video in the form of matrix data, performing processing by means of algorithms consisting of image analysis or "computer vision" algorithms, and acquiring and then providing information regarding temperature variations and / or information regarding the location / distribution of hot spots and / or bands and / or zones on an area of ​​interest of the surface of the brake disc.

[0032] Said acquiring step comprises acquiring, using at least one infrared camera, a plurality of digital images and / or a digital video consisting of a plurality of digital video frames of at least one region of interest of the brake disc to be thermally characterized during said changing dynamic operating conditions, each such digital image and / or digital video frame depicting the temperature detected at each point by means of a color map.

[0033] The aforementioned processing steps include the steps of identifying a region of interest by generating a mask based on an analysis of a plurality of acquired digital images and / or digital video frames, then applying the generated mask to each of the acquired digital images and / or digital video frames to obtain processed matrix data related to the region of interest, then defining an analysis region and performing a thermal analysis on the data related to the defined analysis region to obtain information regarding temperature changes at each point of interest, and finally identifying hot points based on said information regarding temperature changes at each point of interest.

[0034] The providing step may include providing the obtained information in the form of digital images or graphics and / or digital tables and / or digital videos.

[0035] According to one embodiment of the method, said area of ​​interest consists of at least one braking band of a disc.

[0036] According to another embodiment of the method, said regions of interest include two corresponding braking bands on two respective sides of the brake disc.

[0037] In this case, said capturing step includes capturing a plurality of digital images and / or digital videos by at least two infrared cameras, one for each brake band of interest.

[0038] According to another embodiment of the method, said regions of interest include two corresponding braking bands on two respective side surfaces of the brake disc.

[0039] In this case, the acquisition step comprises acquiring a plurality of digital images and / or digital videos using a thermal imaging camera and at least one mirror, the thermal imaging camera directly acquiring digital images and / or digital videos of a portion of interest of one of the brake bands and further acquiring a reflection by the at least one mirror of a digital image and / or digital video of a portion of interest of the other brake band located on an opposite side surface of the brake disc relative to where the thermal imaging camera is located.

[0040] According to one embodiment of the method, said area of ​​interest consists of one or more braking band portions of the brake disc.

[0041] According to an embodiment of the method, the aforementioned operating conditions include at least one braking event or braking test.

[0042] According to an implementation option, the aforementioned operating conditions comprise at least one brake disc bench test, each test including a plurality of brake test events.

[0043] According to an implementation option of the method, the aforementioned operating conditions include a plurality of tests carried out on a brake disc bench, the tests carried out in the series of tests belonging to such a plurality of tests being selectable by an operator supervising the tests.

[0044] According to one embodiment of the method, each test includes a plurality of test braking events, each test braking event having in turn an analysis on each side of the brake disc, each analysis on a brake disc side in turn including an analysis for each frame of digital video captured on such side, and each frame analysis in turn including an analysis on a portion of the frame depicting a respective portion of the region of interest.

[0045] According to one embodiment, some steps of the method are common to all the tests of the aforementioned plurality of tests, while other steps are performed individually for each frame or each of a set of predefined frame parts, which are then expanded to the entire test according to a nesting structure for each test, for each braking action, for each side of the brake side, for each frame and for each given frame part.

[0046] According to one embodiment of the method, the steps of identifying the region of interest by generating a mask and applying the generated mask to each of the acquired digital images and / or digital video frames include computer processing the set of digital images or digital video frames deemed important for analysis.

[0047] Such processing steps include, for each digital image or digital image frame deemed significant, the following steps in sequence: forcing negative temperatures or temperatures below a low temperature threshold (e.g. 10°C) to a low or zero value (typically 0); Applying a filter to the digital image or video frame to reduce noise. Applying Otsu segmentation to the digital image or video frame. Averaging pixel by pixel on the time axis to obtain a respective averaged digital image or averaged digital video frame. Applying Otsu segmentation to the averaged digital image or the averaged digital video frame to obtain a segmented digital image or a segmented digital video frame. Recognizing the edges of the brake disc by an edge recognition algorithm and removing the edges from the segmented digital image or the segmented digital video. Refining the mask by an image cleaning algorithm to finally obtain a processed digital image or a processed digital video.

[0048] According to one embodiment, the obtained digital data corresponding to the processed digital image or the processed digital video is organized in a matrix format, thereby obtaining the processed matrix data related to the region of interest.

[0049] According to an embodiment, the step of obtaining processed matrix data further comprises transforming Cartesian coordinates associated with the indices of the matrices forming the frame into a polar coordinate system.

[0050] According to an implementation option, if each test consists of multiple braking events, the set of digital images or digital video frames considered significant for the analysis consists of the digital images or digital video frames taken from the last N (e.g. 3) braking events of the test, i.e. the events in which the brake discs reach their highest temperature and are therefore more visible against the cool background of the images.

[0051] According to an embodiment, in the recognition step, the aforementioned edge recognition algorithm comprises a Sobel edge detection algorithm.

[0052] According to an embodiment, the image cleaning algorithm comprises a binary image morphology algorithm configured to perform the function of removing dirt and / or small gaps and / or small objects on the image in a refinement step.

[0053] According to one embodiment of the method, said step of defining the analysis region consists of dividing the brake band of the brake disc to be analyzed into three equal subzones: one inner subzone, one central subzone and one outer subzone, each subzone intended to be analyzed separately.

[0054] According to another embodiment of the method, the step of defining the analysis region includes an operator using a command interface to set at least one setting parameter adapted to define an analysis region for each of the two framed brake band portions.

[0055] According to an embodiment, the operator definable configuration parameters are: (i) the angular width of the band under analysis; (ii) Values ​​of the inner and outer radii of the band to define the three subzones to be analyzed (inner, central, and outer subzones).

[0056] According to an embodiment, each pixel of a digital image or video frame is marked to indicate whether it belongs to the analysis region or not, and if so, to which analysis sub-zone it belongs.

[0057] For example, each pixel in each frame included in the analysis region is assigned a number between 1 and 3 that identifies the zone to which it belongs (inner zone, central zone, or outer zone). If the pixel is excluded from the analysis region, it is assigned a 0.

[0058] According to one embodiment of the method, the aforementioned steps of performing a thermal analysis on data associated with a defined analysis area to obtain information regarding temperature changes at each point of interest and identifying thermal hot spots include the following steps: identifying any regions of the analysis region that have a temperature at a predefined percentile higher than the average temperature of the analysis region itself as potential thermal hot spots; calculating the coordinates of the center of gravity for each potential thermal hotspot; Verifying one or more validity requirements for potential thermal hotspots and identifying as thermal hotspots only those potential thermal hotspots that meet said one or more validity requirements.

[0059] According to different implementation options, the validity requirements include one or more or all of the requirements listed below. a. The area of ​​the potential thermal hot spot is greater than the minimum hot spot area threshold. b. The varicenter of a potential thermal hot spot is in a non-cold zone, where a zone is defined as cold if its mean temperature is a given percentile less than the temperature distribution on the brake band side being analyzed, e.g., less than the median temperature of such distribution. c. The area of ​​the potential thermal hotspot is less than the maximum hotspot area threshold. d. The distance between the radial coordinates of the centroids of the two thermal hotspots is greater than the distance threshold, otherwise the two thermal hotspots are considered as one, and the information of the one with the larger area is saved.

[0060] According to an embodiment, the method further comprises the steps of calculating the temperature variation inside each zone and recognizing the presence of a band on the zone considered if said temperature variation is within predetermined limits of temperature variation.

[0061] According to a possible embodiment of the method, the aforementioned information regarding the temperature variation and / or localization / distribution of hot spots and / or bands and / or zones on the area of ​​interest of the surface of the brake disc comprises, for each identified thermal hot spot, information regarding the location (e.g. the centre of gravity or the geometric coordinates of the centre of gravity), and / or area, and / or zone, and the frame in which the hot spot was discovered, and / or metadata identifying the braking event and / or test to which such information refers.

[0062] According to one embodiment of the method, the aforementioned step of providing information regarding the temperature variation and / or location / distribution of hot spots and / or bands and / or zones on an area of ​​interest of the surface of the brake disc comprises displaying the data via a computer interface.

[0063] According to various possible embodiments, such a display step includes displaying the output of the algorithm, and / or displaying (e.g. by means of a line graph) the number of thermal hot spots over time, either globally or split by zones, and / or displaying data regarding the presence or absence of hot bands for each frame, either globally or split by zones, and for each of the two sides of the disk.

[0064] According to an embodiment, the displaying step also includes displaying the reprocessed version of the original digital video frame.

[0065] For example, a digital video frame is re-colored to highlight pixels above the 95th temperature percentile and below the 5th temperature percentile in the same frame that are considered to belong to the same braking band, or pixels corresponding to the centroids of detected thermal hot spots are marked on the digital video frame.

[0066] According to an embodiment, the displaying step further comprises aggregating the reprocessed digital video frames based on the output of the algorithm to recreate the video.

[0067] 1-9, further details of the method are provided below, by way of non-limiting example only, in accordance with certain embodiments of the present invention.

[0068] In this example, we consider a "computer vision" algorithm that is able to automatically extract information about temperature variations and hotspot distribution on the brake disc surface from videos acquired by thermography during dynamic bench tests.

[0069] It should be noted that the term "infrared camera" refers to any electronic means capable of detecting thermal information (e.g., temperature detection by receiving infrared rays) from an object enclosed in a frame and providing as output a digital image or video in which the thermal information is depicted in some way (e.g., in the form of a color map).

[0070] This embodiment is directed to a process for detecting thermal information of a brake disc undergoing test bench testing.

[0071] This process is disclosed below with reference to two steps: acquiring data on a test bench, and processing such data.

[0072] With regard to the data acquisition step, in this embodiment the braking system is tested on a test bench on which an experimental acquisition setup is provided.

[0073] In this example, the temperature change over time of two brake bands of a disk is recorded using an infrared camera and a mirror. The thermograph is positioned so that it directly images part of the brake band and indirectly images part of the brake band on the opposite side of the disk through the reflection of the mirror.

[0074] This recording is done while the disc is rotating, and the recording rate of the infrared thermography is set so that the framed part of the rotating disc is always the same during the entire braking operation. An example of the framing obtained by this recording step is shown in Figure 1.

[0075] Each test provides for the application of a sequence of predefined braking actions in terms of operational parameters (e.g. initial and final speeds, brake system pressure or vehicle deceleration, system temperature at braking start). The relevant data are stored in different frame variables "Frame_00N", one for each video frame and numbered accordingly. Each frame variable stores a float array with certain dimensions (video height) x (video width).

[0076] As previously mentioned, the term frame or digital video frame is used herein to denote a set of data corresponding to the framing of an infrared camera.

[0077] In the embodiment disclosed herein, the frames captured for each braking event are sequentially combined to form a video. The video in matrix format (an example of such a format is MATLAB MAT) is the output of an infrared thermal camera.

[0078] In this example, the data is stored as files on a file system and organized into folders, where each folder represents a different test and contains different video files, and the files are numbered as the position of the corresponding braking event in the sequence of all braking events for a given test.

[0079] The acquisition step is followed by a data processing step, the logical stages of which are shown in Figure 2.

[0080] The analysis units may be different at each step: some operations are performed only once for each test, others for each frame or part thereof separately and then extended over the tests according to a nesting structure of the type: for each test, for each braking action, for each side of the disc, for each frame, for each part (as described above).

[0081] This hierarchy is shown in Figure 3, where, similar to the example described herein, the three parts of the brake band (inner, middle, outer) are separated for each frame.

[0082] Interaction with the user is effected by a software application having a graphic user interface (GUI), for example a specially developed GUI.

[0083] The first step involves the user selecting the test to be analyzed. The software application allows the user to select a folder containing videos related to all braking actions associated with the selected test.

[0084] The next step involves generating a mask corresponding to the analysis zone of interest for this application, namely the braking band of the disc.

[0085] To obtain the mask, the algorithm used in this embodiment of the method performs the following operations (FIG. 4): That is, as input for each test we take the videos of the last three braking operations (i.e. those in which the disc reaches its highest temperature and is therefore more visible against the cool background of the image). In each frame, negative temperatures or temperatures below a certain threshold (eg, 10° C.) are forced to a zero value. To reduce noise, a filter is applied to each frame (Box 1 in Figure 4). Otsu segmentation is applied to each frame (box 2 in FIG. 4). Otsu segmentation (or Otsu's method) is a method known per se for automatically thresholding the histogram of a digital image. Averaging is performed pixel by pixel on the time axis (Box 3 in Figure 4). Apply Otsu segmentation again (box 4 in Figure 4). The Sobel Edge Detection (an algorithm known per se) is used to identify and remove the edges of the disk (box 5 in Figure 4). The mask is refined by binary image morphology (algorithms known per se, e.g. removal of dirt, small gaps, small objects) (box 6 in Figure 4).

[0086] Downstream of this process, an operator can either accept the proposed mask suggested by the algorithm or modify it further based on their needs.

[0087] This comparison is made possible by overlaying the mask in the GUI with a video frame from the test, where the brake band is clearly visible in the background.

[0088] The implementation options allow the operator through a GUI to: 1. Enlarge or reduce the mask. 2. Redefine the contours of the two shapes in the mask based on their intersections with the ellipses identified in the shapes by the RANSAC algorithm. 3. Completely reshape the mask by manipulating the control points of an interpolated Bezier curve.

[0089] Typically, this operation is performed once for the entire test.

[0090] Once the desired mask has been defined, in this embodiment several data processing operations are performed in preparation for the actual temperature analysis. These operations are shown in Figure 5 and are performed for each frame of each braking event during the test.

[0091] First, a mask is applied to each frame of the braking action. Then, each frame is automatically divided into two parts, each relating to and completely containing one side of the brake band. The cuts are made in such a way that only one of the frame's width and length changes. Furthermore, only one position indicator is defined for the cut. From here on, the frames relating to both sides of the brake band are processed in parallel.

[0092] The next step involves transforming the Cartesian coordinates associated with the indices of the matrices forming the frame into a more natural polar coordinate system from the geometry of the object being analyzed.

[0093] Downstream of this step, in this embodiment, through the GUI (see FIG. 6), the user is requested to declare the following parameters for each of the two parts of the framed brake band:

[0094] 1. The user is asked to select a zone where the bandwidth is approximately constant. In practice, due to the dimensions of the test bench, or even the brake caliper alone, it is almost impossible to get a frame of the entire visible angular bandwidth interval.

[0095] 2. Inner and outer radius values ​​of the band. These two values ​​define three zones on the band (inner zone, middle zone, outer zone) where different heat patterns are usually observed. The band is divided into three equal zones by default, but the user can change these values ​​based on the conditions observed in the footage.

[0096] Once these parameters are defined and saved, each pixel in each frame included in the analysis region is assigned an integer number between 1 and 3 that identifies the zone it belongs to (inner zone, central zone, outer zone). If the pixel is excluded from the analysis region, it is assigned the number "0".

[0097] The next step is to analyze the temperature patterns that occurred during the tests. For each of the three zones defined in the brake band of each frame, the following operations are performed (summarized in Figure 7):

[0098] 1. Hotspot (or thermal hotspot) detection. A hotspot is defined as a region that has a temperature that is a certain percentile higher than the temperature of the zone itself.

[0099] 2. For each hotspot, calculate the coordinates of the center of gravity.

[0100] 3. Check validity requirements.

[0101] To eliminate small hot spots such as those caused by holes on the disk, a minimum threshold is set on the area of ​​each hot spot.

[0102] 3b.Only hot spots with their centroid in a non-cold zone are recognized as hot spots. A cold zone is defined as a zone whose mean temperature is a certain percentile lower than the temperature distribution on the analyzed brake band side, for example lower than the median temperature of such distribution. Since hot spots can be defined a priori on any zone, without this device the hottest point of each zone would be detected as a hot spot even if it is in a cold zone, which must be avoided.

[0103] 4. Ensure there are no blended hot spots.

[0104] 4a. Check for angularly blended hotspots.

[0105] The hotspots defined in 1 above may in fact be formed by the junction of several hotspots and may be very extensive. This problem is solved by calibrating the maximum area threshold.

[0106] If the hotspot identified in point 1 above has an area larger than the threshold, the operations of points 1 to 3 above are performed again for that zone using a higher percentile for the search than the percentile selected in point 1. The processed data with the increased percentile are saved and replace the data of the previous process in the angular part of the zone where the too-wide hotspot was detected.

[0107] 4b. Check for radial blended hot spots.

[0108] If the distance between the radial coordinates of the centroids of two hotspots is smaller than a certain threshold, they are considered as one hotspot, and the information about the initial single hotspot with the larger area is stored.

[0109] 5. Calculate the temperature change within each zone and recognize the presence of a band on that zone if the change is within certain predefined limits.

[0110] The output of the operational flow described above contains, for each identified hotspot, information on the location (geometric coordinates of the varicenter), area, zone and frame in which the hotspot was found, as well as identification metadata about the braking operation and the test. These data can be aggregated at different levels and exported in tabular format.

[0111] Within the application it is possible to display the output of the algorithm.

[0112] In particular, once a desired braking action is selected via a drop down menu, various displays are provided to the user.

[0113] For example, a line graph can show the number of hot spots over time, either in total or by zone. Additionally, buttons on the GUI allow data on the presence or absence of hot bands for each frame to be displayed either overall or by zone. Graphs can be tracked for each side of the disc.

[0114] Furthermore, it is also possible to display a re-elaboration of the original frame. In particular, in the embodiment of the method described herein, at least the following options are available: To highlight any errors in the definition (e.g., errors involving the ventilation chamber), the frames considered to belong to the brake band are recolored by highlighting pixels above the 95th temperature percentile and pixels below the 5th temperature percentile in the same frame. Frame in which pixels corresponding to the centroids of hotspots detected by the algorithm are marked.

[0115] The reanalyzed frames based on the output of the algorithm are then aggregated to recreate the video (see Figure 8).

[0116] Finally, it is possible to connect individual displays: for example, a line graph can be made to react to a frame selected on the X-axis to display the corresponding reanalysis frame (see Figure 9). In this way, the user can get visual feedback on the analysis results proposed by the algorithm.

[0117] The graphical outputs, as well as the tabular ones, can be exported and saved by the user / operator.

[0118] Thus, the object of the invention as set out above is fully achieved by the method described above, thanks to the features disclosed in detail above. The advantages and technical problems solved by the method according to the invention have already been described above with reference to various features and aspects of the method.

[0119] In particular, the above-mentioned solution allows an accurate, rapid, objective, reproducible and inexpensive analysis of data provided by an infrared camera framing a disc brake under dynamic conditions.

[0120] To meet fortuitous needs, those skilled in the art can modify and adapt the above-described method embodiments or substitute other functionally equivalent elements without departing from the scope of the following claims. Each feature described as belonging to a possible embodiment can be achieved independently of the other embodiments described.

Claims

1. 1. A method for automatically detecting, by computer processing, information about temperature changes on the surface of a brake disc and / or information about the location / distribution of hot spots and / or bands and / or zones on the surface of a brake disc under dynamic operating conditions, the method comprising: acquiring, by at least one thermal imaging camera, a digital video comprising a plurality of digital images and / or a plurality of digital video frames of at least one area of ​​interest of the brake disc to be thermally characterized, wherein each of the plurality of digital images and / or the plurality of digital video frames depicts the temperature detected at each point by a color map, and wherein the acquiring is performed during changing dynamic operating conditions; obtaining a representation of each of said plurality of digital images and / or said plurality of digital video frames in the form of matrix data; processing said plurality of digital images or said plurality of digital video frames by algorithms, including image analysis algorithms or computer vision algorithms, to obtain information about said temperature variations and / or information about the location / distribution of hot spots and / or bands and / or zones on said area of ​​interest of the surface of said brake disc; providing said acquired information in the form of a digital image or graphic and / or a digital table and / or a digital video; The processing step includes: identifying the region of interest by generating a mask based on analysis of the plurality of acquired digital images and / or the plurality of acquired digital video frames; applying the generated mask to each of the plurality of acquired digital images and / or the plurality of acquired digital video frames to obtain processed matrix data related to the region of interest; A step of defining an analysis domain; performing a thermal analysis on data associated with the defined analysis region to obtain information about the temperature change at each point of interest; and identifying hot spots based on information about the temperature changes at each of the points of interest.

2. The method of claim 1 , wherein the region of interest includes at least one braking band of the brake disc.

3. the region of interest includes two corresponding brake bands on two respective sides of the brake disc, 2. The method of claim 1, wherein the acquiring step includes acquiring the plurality of digital images and / or the plurality of digital video frames with at least two of the thermal imaging cameras, one for each brake band of interest.

4. the region of interest includes two corresponding brake bands on two respective side surfaces of the brake disc, the acquiring step includes acquiring the plurality of digital images and / or the plurality of digital video frames with a thermal imaging camera and at least one mirror; the thermal imaging camera directly acquiring the plurality of digital images and / or the digital video frames of a portion of interest of one of the brake bands; 2. The method of claim 1, further comprising acquiring a reflection provided by at least one mirror of the digital image and / or the digital video frame of the portion of interest of another brake band located on an opposite side of the brake disc with respect to where the thermal imaging camera is located.

5. The method of claim 1 , wherein the region of interest includes portions of one or more brake bands of the brake disc.

6. The method of any one of claims 1 to 5, wherein the operating conditions include at least one braking event or braking test.

7. The method of claim 6 , wherein the operating conditions include at least one brake disc bench test, the brake disc bench test including a plurality of brake test events.

8. The operating conditions include a plurality of tests performed on a brake disc bench; 8. The method of claim 7, wherein tests within the plurality of tests to be performed in a sequence of tests are selectable by an operator performing the plurality of tests.

9. each of the plurality of tests includes a plurality of test braking events; the test braking event includes an analysis of each side of the brake disc; the analysis of each side of the brake disc includes analysis of each frame of the plurality of digital video frames taken of each side; The method of claim 8 , wherein the analysis for each frame includes analysis for a portion of the plurality of digital video frames depicting a respective portion of the region of interest.

10. some steps of the method are common to all tests of the plurality of tests; 10. The method of claim 9, wherein the other steps of the method are performed individually for each frame or set of predetermined frame portions and are extended throughout the tests in a nested structure for each test, each braking action, each side of the brake disc, each frame, and each frame portion.

11. wherein the step of identifying the region of interest by generating a mask and applying the generated mask to each of the plurality of acquired digital images and / or plurality of digital video frames comprises processing, by a computer, a set of digital images or digital video frames deemed significant for analysis; For the digital images or digital video frames deemed significant, the processing step comprises: setting negative temperatures or temperatures below a predetermined low temperature threshold to a low or zero value; applying a filter to the digital images or digital video frames deemed significant to reduce noise; applying Otsu segmentation to said digital images or said digital video frames deemed significant; averaging pixel by pixel over a time axis to obtain an averaged digital image or an averaged digital video frame; applying Otsu segmentation to the averaged digital image or the averaged digital video frame to obtain a segmented digital image or a segmented digital video frame; Recognizing edges of the brake disc by an edge recognition algorithm and removing said edges from the segmented digital image or the segmented digital video; The method according to any one of claims 1 to 5, comprising a refining step, in which the mask is refined and purified by an image cleaning algorithm to finally obtain a processed digital image or a processed digital video.

12. 12. The method of claim 11, wherein the step of obtaining processed matrix data further comprises converting Cartesian coordinates associated with matrix indices forming the digital video frame to a polar coordinate system.

13. where each of the plurality of tests comprises a plurality of braking events, the set of digital images or digital video frames deemed significant for analysis comprises digital images or digital video frames taken from the last N braking events of the test, the N events in which the brake discs reached their highest temperature and were therefore more visible against the cooler background of the images; the edge recognition algorithm includes a Sobel edge detection algorithm in the recognition step; The method of claim 11 , wherein the image cleaning algorithm comprises a binary image morphology algorithm configured to perform the function of removing dirt and / or small gaps and / or small objects on the image in the refining step.

14. The step of defining the analysis domain includes: dividing the braking band of the brake disc to be analyzed into three equal subzones: one inner subzone, one central subzone, and one outer subzone, each subzone intended to be analyzed; or an operator using a command interface to set at least one setting parameter adapted to define the analysis region for each of the two portions of the digital video framed brake band, the at least one setting parameter comprising: (i) the angular width of the band of interest of the analysis; (ii) values ​​of inner and outer radii on the brake band adapted to define three subzones: one inner subzone, one central subzone and one outer subzone, each subzone intended to be analyzed; 6. A method according to any one of claims 1 to 5, wherein each pixel of the digital image or digital video frame is marked to indicate whether it belongs to the analysis region or not, and if so, to which analysis subzone it belongs.

15. performing a thermal analysis on data related to the defined analysis area to obtain information about the temperature change at each point of interest; and identifying hot spots based on the information about the temperature change at each point of interest; identifying any region of the analysis region having a temperature that is a predetermined percentile higher than the average temperature of the analysis region as a potential hot spot; calculating the coordinates of the center of gravity for each of said potential hotspots; verifying one or more validity requirements of the potential hotspots and identifying as hotspots only those potential hotspots that satisfy the one or more validity requirements; The one or more validity requirements: (a) the area of ​​the potential hotspot is greater than a minimum hotspot area threshold; (b) the centroid of the potential hot spot is in a non-cold zone other than an area defined as cold if the average temperature is less than a predetermined percentile of the temperature distribution of the brake band being analyzed or is less than the median temperature of the distribution; (c) the area of ​​the potential hotspot is less than a maximum hotspot area threshold; (d) the distance between the radial coordinates of the centroids of two hotspots is greater than a predetermined distance threshold; otherwise, the two hotspots are considered as one, and information on the hotspot with the larger area is saved.

16. calculating the temperature change in each region; and recognizing the presence of a band on said zone if said temperature change is within a preset temperature change limit.

17. A method according to any one of claims 1 to 5, wherein the information relating to the temperature changes and / or the location / distribution of hotspots and / or bands and / or zones on the area of ​​interest on the surface of the brake disc includes, for each identified hotspot, information about the location and / or area and / or band and the digital video frame in which the hotspot was found, and / or metadata identifying the braking event and / or test to which the information relates.

18. - providing information about the temperature variation in the area of ​​interest on the surface of the brake disc and / or about the location / distribution of hot spots and / or bands and / or areas, displaying the hotspot recognition output; and / or Displaying the total number of hotspots or the number per area over time; and / or Displaying data on the presence or absence of hot bands for each frame, either globally or divided into regions, for each of the two sides of the disc; and / or Displaying a re-elaborate version of the original digital video frame; and / or A method according to any preceding claim, comprising aggregating re-refined digital video frames based on the output of said algorithm to recreate the video.