Classifications

• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T7/00—Image analysis
  • G06T7/0002—Inspection of images, e.g. flaw detection
  • G06T7/0004—Industrial image inspection
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
  • G06V10/00—Arrangements for image or video recognition or understanding
  • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
  • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
  • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
  • G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
  • G06V10/00—Arrangements for image or video recognition or understanding
  • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
  • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T2207/00—Indexing scheme for image analysis or image enhancement
  • G06T2207/20—Special algorithmic details
  • G06T2207/20021—Dividing image into blocks, subimages or windows
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T2207/00—Indexing scheme for image analysis or image enhancement
  • G06T2207/20—Special algorithmic details
  • G06T2207/20081—Training; Learning
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T2207/00—Indexing scheme for image analysis or image enhancement
  • G06T2207/20—Special algorithmic details
  • G06T2207/20084—Artificial neural networks [ANN]
• • G—PHYSICS
  • G06—COMPUTING OR CALCULATING; COUNTING
  • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
  • G06T2207/00—Indexing scheme for image analysis or image enhancement
  • G06T2207/30—Subject of image; Context of image processing
  • G06T2207/30108—Industrial image inspection
  • G06T2207/30164—Workpiece; Machine component

Definitions

• TITLE Process for detecting defects on an aeronautical part
• the technical field of the invention is that of the detection of defects, and more particularly that of the automatic detection of defects on an aeronautical part.
• each image is divided into several sub-images and the check is carried out for each sub-image.
• Two strategies are then possible to find the defects, either a classification strategy consisting in indicating whether the sub-image has a defect, or a semantic segmentation strategy consisting in indicating on each sub-image, the pixels included in a defect zone .
• the invention offers a solution to the problems mentioned above, by making it possible to detect the presence of defects on an aeronautical part via an ensemble method using the optimal epochs of the training of an artificial neural network making it possible to limit the number of false alarms.
• a first aspect of the invention relates to a fault detection method for an aeronautical part from at least one image to be inspected of the aeronautical part subdivided into a plurality of sub-images, the method comprising the following steps :
• N being an integer greater than or equal to 1
• a “business criterion” is understood to mean a criterion for selecting curves which is based on the knowledge of the user of the method according to the invention and/or in depending on the application concerned for the method according to the invention. It may, for example, be a criterion based on the area under the curves, a detection density in the curves, a detection pattern such as a high detection density then low detection density, regularity of the curves, etc.
• the term "defect” means a discontinuity in the properties of the material making up a part or an inspected object, in this case the aeronautical part. This discontinuity results from an anomaly present in the material. This anomaly can have various origins and be of varied nature. These anomalies are mainly the consequence of hazards that occur during the manufacture of the part. These anomalies also occur quite frequently during the use of the part or its handling: the material may, for example, have been weakened during the manufacturing process and its use, generating strong local stresses at the level of the weakened zone, or following an impact, generates a defect. The term “defect” therefore covers all forms of anomalies that the material may suffer: material defect, inclusion, crack, porosity, corrosion, alteration of the properties of the material, etc.
• Fault detection curve as a function of false alarms is understood to mean a set of points connected with the number of false alarms on the abscissa (which may range from 0 to infinity) and the number of faults correctly detected on the ordinate (which may go from 0 to the total number of faults to be detected, which can be zero in some cases).
• the possible presence of defects is detected on each sub-image composing an image of the aeronautical part, by merging the results obtained using an artificial neural network taken at several periods of 'training. These epochs were pre-selected to detect the approximate location of each fault while limiting the number of false alarms.
• a results matrix is reconstituted for each image of a validation database from the results provided by the artificial neural network trained during the training epochs for each sub-image composing the image, which makes it possible to obtain a view of the defects on the scale of the image and thus to remove certain biases introduced by the division into sub-images.
• a curve representing the number of detections as a function of the number of false alarms is then determined for each epoch. It is then possible to select a set of curves according to a criterion job. This makes it possible to select the curves to meet performance criteria in order to select the optimal training epochs for the detection of faults by minimizing the number of false alarms.
• the neural network is configured to learn and predict on sub-images, then the data produced at the output of the neural network is used to evaluate the performance of the neural network on the images from which the sub-images come. .
• the performances on the images are therefore not calculated directly but indirectly by the calculation of the performances on the sub-images.
• the method according to the invention may have one or more additional characteristics among the following, considered individually or according to all technically possible combinations.
• the supervised training is carried out, for the plurality of epochs, from a training database, the number of training epochs of the plurality of epochs of the training being determined by a stopping criterion, each training epoch among the plurality of training epochs being associated with a number, the training database comprising a plurality of sub-images for each image of a plurality of training images of aeronautical parts, each sub-image being associated with defect information for each pixel of the sub-image.
• a “stopping criterion” is called a criterion which makes it possible to provide information on the condition or conditions for stopping the learning process of the artificial neural network. It may for example be a predefined maximum number of training epochs, a stopping threshold to be reached for the cost function of the artificial neural network, the detection of over-learning, etc.
• the step of creating a default probability matrix associated with each training epoch and associated with each image of the validation set is done according to the following sub-steps:
• the step of determining a fault detection curve according to the false alarms associated with a training period is done according to the following sub-steps:
• Each result matrix obtained is thresholded with several thresholds to take into account a diversity of operating points for each epoch, and in each thresholded result matrix, the sets of non-damaged adjacent pixels are compared with the corresponding defect information. to calculate the number of detections and the number of false alarms, independent of the number of pixels included in the detection and the size of the false alarm.
• the control result of a sub-image is a matrix, each coefficient of which corresponds to a probability of default for each pixel of the sub-image.
• the artificial neural network is configured to provide, from a sub-image, a probability of defect per pixel of the sub-image, each sub-image of the database of training and each sub-image of the validation database being associated with defect information per pixel of the sub-image.
• the artificial neural network is trained and tested to perform a semantic segmentation task.
• the artificial neural network is configured to provide, from a sub-image, a probability of default for all the pixels of the sub-image, each sub-image of the base of training data and each sub-image of the validation database being associated with defect information for all the pixels of the sub-image.
• the artificial neural network is trained and tested to perform a classification task.
• the step of determining the fault detection curve further comprises a step of correction by deletion of each point corresponding to a threshold for which the number of false alarms associated with the threshold is less than the number of false alarms associated with the immediately following threshold in the predefined threshold interval following the Plot step, the deletion step being restarted at each point deleted from the lowest remaining threshold.
• the curves are corrected for an aberrant behavior of the number of false alarms when the detection threshold applied does not make it possible to correctly discern the regions of the images with a detected defect and those with a false alarm.
• this makes it possible to correct the curves when the number of false alarms decreases while the value of the threshold decreases.
• the corrected curve is normalized by adding an initial point associated with a zero number of detections and a zero number of false alarms and adding a final point associated with the maximum number of detections of the corrected curve and a predefined maximum number of false alarms, depending on the correction step.
• the thresholded default probability matrix is smoothed.
• the business criterion for determining a set of the N best training epochs of the plurality of training epochs as a function of the fault detection curves is the area under the curve .
• the detection of a defect for an image to be checked is done by:
• each pixel of an image is associated with a position in the image and the step of creating the matrix of fault probabilities for an image is also carried out from the associated position. at each pixel of the image.
• the defect probability matrix corresponds to the image of the validation database in which the color or gray level information associated with each pixel is replaced by the pixel defect probability.
• a fusion is carried out, for each pixel of the image associated with a plurality of fault probabilities, of the plurality of probabilities defect associated with the pixel.
• the pixel fault probability associated with the pixel takes into account the fault probability obtained for the pixel for each sub-image.
• the fault probabilities for each sub-image of the image to be checked obtained at the end of the step are merged in order to obtain fault probabilities for the image to be checked.
• the business criterion for the selection of a set of curves is a predetermined number of curves corresponding to the number of curves whose area under the curve is the highest.
• the artificial neural network is configured to provide, from a sub-image, a probability of default for all the pixels of the sub-image, each sub-image of the base of training data and each sub-image of the validation database being associated with defect information for all the pixels of the sub-image.
• the artificial neural network is trained and tested to perform a classification task.
• all the coefficients of the result matrix associated with the sub-image will be equal to the default probability provided by the network for this sub-image.
• the merging of the fault probabilities obtained for a pixel is carried out by averaging the fault probabilities obtained for the pixel.
• Another aspect of the invention relates to a computer program product comprising instructions which, when the program is executed on a computer, lead the latter to implement the steps of the method according to the invention.
• Figure 1 shows a schematic representation of an aeronautical part image used by a method according to the invention.
• Figure 2 is a block diagram illustrating the sequence of steps of the method according to the invention.
• the invention relates to a method for detecting defects on an aeronautical part to be inspected.
• the detection of faults is automatic, that is to say that it does not require any human intervention during the final use phase of the system, and is carried out from one or more images of the part. aeronautics, for example taken with different viewing angles.
• Each image of the aeronautical part is acquired beforehand by a camera and can be in color or in grayscale.
• Figure 1 shows a schematic representation of the image 200 of the aeronautical part.
• the image 200 of the aeronautical part comprises a set of pixels 2010 and is subdivided into a plurality of sub-images 201 each comprising a subset of pixels 2010 of the set of pixels 2010.
• Figure 1 is shown an image 200 comprising a set of 36 pixels 2010.
• the plurality of sub-images 201 may or may not have areas of overlap between them.
• FIG. 1 two sub-images 201 each comprising a subset of 9 pixels and overlapping on a pixel 2010 shown hatched.
• the aeronautical part is preferably of complex shape, such as an engine blade, different types of blade root attachment, an aircraft fairing, etc.
• Figure 2 is a block diagram illustrating the sequence of steps of the method 100 according to the invention.
• a first step 101 of the method 100 according to the invention consists in training in a supervised manner an artificial neural network on a training database, to obtain an artificial neural network capable of providing from a sub -image 201 , a default probability for each pixel 2010 of sub-image 201.
• the defect probability associated with a pixel 2010 expresses the probability that the pixel 2010 belongs to a defect and is between 0 and 1, or between 0 and 100% if it is expressed as a percentage.
• Supervised training otherwise called supervised learning, makes it possible to train an artificial neural network for a predefined task, by updating its parameters so as to minimize a cost function corresponding to the error between the data of output provided by the artificial neural network and the real output datum, i.e. what the artificial neural network should output to fulfill the predefined task on a certain input datum.
• the hyperparameters are, for example, the number of layers of the network, the number of neurons per layer, the cost function to be optimized, the predefined parameters of the optimization algorithm, etc.
• the parameters are, for example, the weights of the neurons.
• a training database therefore comprises input data, each associated with a real output data.
• the training database comprises a plurality of images 200 of aeronautical training parts, each subdivided into a plurality of sub-images. images 201 , each sub-image 201 being associated with defect information for each pixel 2010 of the sub-image 201 .
• the input data are the sub-images 201 and the real output data are the fault information for each pixel 2010 of the sub-image.
• the defect information associated with a pixel 2010 is equal to 0 if the pixel 2010 does not belong to a defect and to 1 or 100% if the pixel 2010 belongs to a defect.
• the aeronautical drive parts are of the same type as the aeronautical part to be checked, that is to say that if the aeronautical part to be checked is an aircraft engine blade, each aeronautical drive part is also an aircraft engine blade.
• the training database comprises for example 200 images 200.
• the artificial neural network can be trained for a classification task and then provides a single default probability for all the pixels 2010 of the sub-image 201.
• Each sub-image 201 of the database of training is then associated with fault information for all the pixels 2010 of the sub-image 201 .
• the artificial neural network trained for the classification task is for example a ResNet residual artificial neural network.
• the artificial neural network can also be trained for a semantic segmentation task and then provides a probability of failure per pixel 2010 of the sub-image 201.
• Each sub-image 201 of the training database is then associated with defect information per pixel 2010 of the sub-image 201.
• the artificial neural network trained for the semantic segmentation task is for example the LinkNet, U-Net or FDNN artificial neural network.
• the training of the artificial neural network is carried out by carrying out a certain number of epochs on the training data base, for example until a predefined stopping criterion is satisfied.
• a stopping criterion can be a predefined number of epochs, for example 200 epochs.
• Another example would be to train the network for a given time, for example a week.
• a another example would be to train the network until over-training occurs on the validation set (e.g. the cost function continues to decrease on the training set, but begins to increase for a predefined number of epochs on the validation set).
• the term "the training of an artificial neural network is carried out by carrying out N epochs on a database comprising M data", the following iterative process: during the first epoch, the M data of the training base (taken in order or randomly), during the second epoch, the M data of the training base are provided (taken in the same order or randomly with respect to the previous epoch), ... , during the n-th epoch, the M data of the training base are provided (taken in the same order or randomly with respect to the previous epoch).
• the M data of the database are provided N times as input to the artificial neural network during its training, each of the M data being provided only once during each of the epochs.
• the plurality of epoch numbers includes, for example, each epoch number carried out up to the stop condition, that is to say that if the stop condition is triggered at epoch number 300 , the plurality of epoch numbers comprises for example each epoch number between 1 and 300.
• the hyperparameters and parameters of the artificial neural network have for example been saved at the end of each training epoch during the first step 101.
• a second step 102 of the method 100 according to the invention is carried out for each training period, and for each image of a validation set V1, each image 200 of the validation set V1 being subdivided into a plurality of sub - 201 pictures.
• validation set is meant a database of validation images 200, the validation images 200 being used for the validation of learning during the second step 102 of the method 100 according to the invention.
• the validation database comprises a plurality of aeronautical part validation images 200, each subdivided into a plurality of sub-images. images 201 , each validation sub-image 201 being associated with fault information for each pixel 2010 of the sub-image 201 .
• the validation aeronautical parts are of the same type as the aeronautical part to be checked.
• the validation database comprises for example 200 validation images 200 .
• each sub-image 201 of the validation image database is then associated with defect information for all the pixels 2010 of subpicture 201 .
• each sub-image 201 of the validation image database is then associated with defect information per pixel 2010 of the sub-image. -picture 201.
• the second step 102 of the method 100 consists, for each training epoch and for each image 200 of the validation set V1, in creating a default probability matrix from a fusion of the probabilities of faults determined for each pixel 2010 of each sub-image 201 of the image 200 of the validation set, the fault probabilities being determined during a sub-step 1021 by applying the artificial neural network to each sub-image 201 of the picture
• the artificial neural network being applied with parameters associated with said training epoch.
• each pixel 2010 of each validation image 200 of the validation image database is for example associated with at least one fault probability provided by the trained artificial neural network.
• a pixel 2010 can be associated with a plurality of fault probabilities if the pixel 2010 belongs to an area of overlap between several sub-images 201 , as is the case for the hatched pixel 2010 in FIG. 1 .
• the defect probability matrix associates with each pixel 2010 of the validation image 200 considered, a pixel defect probability obtained from each defect probability associated with the pixel 2010 at the end of the second step 102 for the number of epochs considered.
• the pixel fault probability is equal to the single associated fault probability.
• the second step comprises for example a sub-step 1022 consisting in obtaining the pixel fault probability by merging the plurality default probabilities associated with pixel 2010.
• the merging is for example performed by calculating the average of the plurality of fault probabilities associated with the pixel 2010.
• the fault probability matrix corresponds to the validation image 200 in which for each pixel 2010, the information of color or grayscale associated with pixel 2010, is replaced by the pixel defect probability, i.e. the defect probability matrix is arranged according to the position associated with each pixel 2010 in the frame 200 validation.
• the defect probability matrix can therefore be likened to a gray level image in which a pixel defect probability equal to 0 is represented in black and a pixel defect probability equal to 1 or 100% is shown in white.
• each validation image 200 of the validation database is associated with a default probability matrix for the training period considered.
• a third step 103 of the method 100 according to the invention consists, for each training epoch, in determining from the probability matrices calculated in step 102 associated with said epoch, a fault detection curve as a function false alarms.
• the step 103 of determining a detection curve, for each training epoch, can comprise a plurality of sub-steps.
• the predefined threshold interval is equal to [0; 1] or [0; 100] if default probabilities are expressed as a percentage.
• the plurality of thresholds is for example chosen from the predefined interval of thresholds so that, for example, the plurality of thresholds is regularly spaced. If the predefined threshold interval is equal to [0; 1], the plurality of thresholds is for example chosen from 0.1 to 0.9 in steps of 0.1, and therefore comprises 9 thresholds. It is also possible to define the plurality of thresholds with intervals of variable size, in particular if the curve varies greatly around certain values, for example to be more representative of the criticality of the values. In addition, the plurality of thresholds can be defined according to business criteria.
• a first sub-step 1031 of step 103 of the method 100 according to the invention consists, for each matrix of default probabilities obtained at the end of the second step 102, in thresholding the matrix of default probabilities considered by the threshold considered, i.e. replacing each pixel defect probability of the defect probability matrix less than or equal to the threshold by 0 and each pixel defect probability of the defect probability matrix strictly greater at threshold by 1 .
• the thresholded defect probability matrix can therefore be likened to a binary image in which a pixel defect probability equal to 0 is represented in black and a pixel defect probability equal to 1 or 100% is represented in black. white.
• each image 200 of the validation database is associated with a thresholded default probability matrix for the number of epochs considered.
• the method 100 according to the invention may comprise a step consisting, for each thresholded default probability matrix obtained at the end of the first sub-step 1031 , in smoothing the considered thresholded default probability matrix, for example by morphological opening or closing operations.
• a second sub-step 1032 of step 103 of the method 100 according to the invention is then carried out.
• This sub-step 1032 consists in incrementing a number of false alarms associated with the threshold considered, the number of false alarms associated with the threshold being initially zero, each time a condition C1 is fulfilled, by applying it to each set of adjacent pixels 2010 associated with a non-zero pixel defect probability after the thresholding carried out in the first sub-step 1031 and possibly the subsequent smoothing.
• Condition C1 is verified if each pixel 2010 of the considered set of adjacent pixels 2010 associated with a non-zero pixel defect probability after the thresholding carried out in the first sub-step 1031 and possibly the smoothing carried out thereafter is associated with zero defect information in the validation database.
• Each set of adjacent pixels 2010 must for example have a connectivity of at least 4. Thus, if for example in the entire validation base, 200 sets of adjacent pixels 2010 are obtained associated with a non-zero pixel defect probability after the thresholding carried out at the first sub-step 1031 and possibly the smoothing performed afterwards and condition C1 is fulfilled 100 times, at the end of the second sub-step 1032 the number of false alarms associated with the considered threshold will be 100.
• a third sub-step 1033 of step 103 of the method according to the invention is then performed.
• This sub-step 1033 consists in incrementing a number of detections associated with the considered threshold, the number of detections associated with the threshold being initially zero, each time a condition C2 is fulfilled, by applying it to each set of associated adjacent pixels 2010 to non-zero fault information in the validation database.
• Condition C2 is verified if at least one pixel of the considered set of adjacent pixels associated with non-zero defect information in the validation database is also associated with a non-zero pixel defect probability after the thresholding performed at the first sub-step 1031 and possibly the smoothing carried out afterwards.
• Each set of adjacent 2010 pixels must for example be of at least 4 connectivity.
• the number of detections associated with the considered threshold will be 100.
• each threshold is associated with a number of detections and with a number of false alarms having been obtained from each default probability matrix created for the period number considered.
• a fourth sub-step 1034 of step 103 of the method 100 according to the invention consists in plotting, for the number of epochs considered, a curve representing the number of detections as a function of the number of false alarms.
• FIG. 3 represents a first curve 1081 and a second curve 1081 obtained at the end of the fourth sub-step 1034 for respectively a first training epoch number among the plurality of training epochs and a second training epoch number among the plurality of training epochs.
• the curve 1081 comprises a point 1082 per threshold considered, corresponding to the number of detections and the number of false alarms associated with the threshold obtained at the end of the second sub-step 1032 and the third sub-step 1033 of the step 103 of the method 100 according to the invention.
• the first curve 1081 and the second curve 1081 each comprise 8 points 1082.
• a fifth sub-step 1035 of step 103 of the method 100 according to the invention consists in correcting the curve 1081 obtained in the fourth sub-step 1034 for the number of epochs considered.
• the correction is made by deleting each point 1082 corresponding to a threshold for which the number of associated false alarms is less than the number of false alarms associated with the immediately following threshold in the predefined threshold interval (the nominal behavior is that the number of false alarms decreases when the threshold value increases).
• the deletion process must thus be done iteratively, starting with the point corresponding to the lowest threshold.
• the deletion process is restarted from the lowest remaining threshold as soon as a point has been deleted, until no more points need to be deleted and considering the immediately following threshold among the undeleted thresholds.
• the plurality of thresholds can comprise a first threshold equal to 0.3 associated with a number F1 false alarms, a second threshold equal to 0.6 associated with a number F2 false alarms, a third threshold equal to 0.7 associated with a number F3 false alarms, a fourth threshold equal to 0.8 associated with a number F4 false alarms, and a fifth threshold equal to 0.9 associated with a number F5 false alarms such that F4>F2>F1>F3>F5. Since the point 1082 associated with the first threshold has fewer false alarms than the point associated with the second threshold, the point associated with the first threshold is deleted. Thus, since a point has been deleted, the lowest remaining threshold point is chosen to begin the deletion process again.
• the lowest remaining threshold point is the point associated with the second threshold.
• the point associated with the second threshold is associated with more false alarms than the point associated with the third threshold, it is not deleted.
• the point associated with the third threshold is associated with fewer false alarms than the point associated with the fourth threshold, the point associated with the third threshold is therefore deleted.
• the lowest remaining threshold point is chosen to begin the deletion process again.
• the lowest remaining threshold point is the point associated with the second threshold.
• the point associated with the second threshold is associated with fewer false alarms than the point associated with the fourth threshold, the point associated with the second threshold is therefore deleted.
• the lowest remaining threshold point is chosen to begin the deletion process again. In this case, it is the point associated with the fourth threshold which is associated with more false alarms than the point associated with the fifth threshold.
• the fourth threshold is thus not deleted.
• the fifth threshold being the largest threshold, it has no point associated with an immediately higher threshold, so it is not deleted.
• a sixth sub-step 1036 of step 103 of the method 100 according to the invention consists in normalizing the corrected curve 1081 obtained in the fifth sub-step 1035 for the number of epochs considered. .
• FIG. 4 represents the first curve 1081 and the second curve 1081 of FIG. 3, obtained at the end of the sixth sub-step 1036.
• the normalization is carried out by adding to the curve 1081 an initial point 1083 associated with a zero number of detections and a zero number of false alarms and an end point 1084 associated with the maximum number of detections of the corrected curve 1081 obtained at the fifth sub-step 1035 and a predefined maximum number of false alarms.
• the predefined maximum number of false alarms is for example equal to the maximum number of false alarms by comparing the curves obtained in the fourth step 1034 or in the fifth sub-step 1035 for each number of epochs considered.
• a normalized curve 1081 is obtained for each epoch number of the plurality of epoch numbers.
• a fourth step 104 of the method 100 according to the invention consists in determining a predefined number N of the best training periods among the plurality of training periods as a function of the fault detection curves 1081 and as a function of a business criterion.
• the determination of the N best epochs can be carried out by selecting N fault detection curves 1081 as a function of the business criterion.
• the selection can be performed for example automatically via a selection algorithm or be established by an operator.
• the business criterion for making the selection is, for example, based on the knowledge of the operator and/or depending on the application concerned.
• the criterion can be based on the area under the curves, a detection density in the curves, a detection pattern such as a high detection density then a low detection density, the regularity of the curves, etc
• a sub-step 1041 of the determination step 104 consists in calculating the area under each normalized curve 1081 obtained at the end of the sixth sub-step 1036 of the step 103 of the method 100 according to the invention and to select a predefined number N of curves 1081 corresponding to the curves 1081 having the N highest calculated areas.
• N is for example between 2 and 50.
• the N best epochs corresponding to the N selected curves have been determined and selected.
• determination of an epoch is meant the selection of the state of the network at this epoch.
• One thus obtains N sets of known parameters associated with the N epochs which can be loaded into the network a posteriori.
• a fifth step 105 of the method 100 according to the invention is carried out for each training epoch of the set of the number N of best training epochs.
• the fifth step 105 consists for each training epoch of the set of the number N of best epochs, in an application 105, on each sub-image 201 of the image 200 to be controlled, of the artificial neural network with the parameters associated with said epoch, to obtain a result matrix of the controlled sub-image associated with said training epoch.
• Each coefficient of the result matrix of a controlled sub-image can correspond to a probability of defect of a pixel of the sub-image 201 .
• each pixel 2010 of the sub-image 201 is associated with a plurality of fault probabilities comprising at least one fault probability per selected epoch number.
• a sixth step 106 of the method 100 according to the invention consists, for each sub-image 201 of the image to be checked, in merging result matrices of the checked sub-images, each result matrix of a sub-image controlled coming from the fifth step 105.
• the fusion is for example carried out by calculating the mean of the coefficients of each result matrix corresponding to the controlled sub-images, the result of the fusion being a matrix in which each coefficient is equal to or depends on the mean of each coefficient of the same position of each result matrix of a controlled sub-image.
• a defect prediction is obtained for each sub-image 201 of the image 200 to be checked, corresponding to the probability that a defect is found in the sub-image 201 .
• This probability is either given pixel per pixel (in the case of semantic segmentation), or given for the entire sub-image (in the case of classification).
• a seventh optional step 107 consists in reconstituting a result matrix of the controlled image 200 by merging the result matrices of the sub-images 201 obtained at the end of the sixth step 106.
• the fault prediction is obtained for the image 200 to be checked, corresponding to the fact that a defect is found in the image 200. This probability is given pixel by pixel.

Landscapes

• Engineering & Computer Science (AREA)
• Theoretical Computer Science (AREA)
• Computer Vision & Pattern Recognition (AREA)
• General Physics & Mathematics (AREA)
• Physics & Mathematics (AREA)
• Evolutionary Computation (AREA)
• Computing Systems (AREA)
• General Health & Medical Sciences (AREA)
• Medical Informatics (AREA)
• Software Systems (AREA)
• Databases & Information Systems (AREA)
• Artificial Intelligence (AREA)
• Health & Medical Sciences (AREA)
• Multimedia (AREA)
• Quality & Reliability (AREA)
• Image Analysis (AREA)
• Image Processing (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=81580886

Family Applications (1)

Country Status (5)

Families Citing this family (2)

• 2022
  • 2022-02-16 FR FR2201348A patent/FR3132779B1/fr active Active
• 2023
  • 2023-01-24 WO PCT/FR2023/050093 patent/WO2023156721A1/fr not_active Ceased
  • 2023-01-24 EP EP23707128.7A patent/EP4479931A1/de active Pending
  • 2023-01-24 CN CN202380022223.5A patent/CN118715542A/zh active Pending
  • 2023-01-24 US US18/838,297 patent/US20250157020A1/en active Pending

Also Published As

Similar Documents

Legal Events