FIELD: control of technological processes.
SUBSTANCE: according to the method for detecting extended defects, a module consisting of a linear scanning chamber and at least one linear directional light source is used, the resulting frame is divided into blocks and p blocks are obtained on the horizontal and q blocks on the vertical, where each block with number i, j, 0 ≤ i < p, 0 ≤ j < q, numerical values of its properties are generated. A modified halftone morphology operation is applied to the image, the central pixel of the morphological structuring element is replaced with the r-th element of the order statistics in the region covered by the structuring element, where r is determined by the detected defect image minimum number of pixels, and the brightness distribution law of the image pixels after the halftone morphology is adaptively reduced to the normal law with a zero average. For each block with a number i, j, calculating the mathematical expectation mi, j and the standard deviation σi, j and when reducing the brightness of each pixel, bilinear interpolation of values of mathematical expectation and mean-square deviations in blocks adjacent to said pixel is applied, filtering the image with the reduced brightness distribution law using at least two spatial quasi-matched filters with different aperture sizes, comparing for each pixel the filtering result of each quasi-matched filter with a threshold and, if this threshold is exceeded, decision is made on the presence of a material texture defect in a given pixel, and the threshold processing results are combined according to the disjunction rule.
EFFECT: high probability of correct detection of small-size defects.
1 cl, 10 dwg
Title | Year | Author | Number |
---|---|---|---|
METHOD FOR AUTOMATIC RECOGNITION OF TRACES OF FIREARMS ON PICTURE OF LATERAL SURFACE OF BULLET (OR CASE) | 2009 |
|
RU2414748C1 |
NODE FOR MONITORING CONDITION OF STEEL ROPES OF LOAD-LIFTING MECHANISMS FOR VARIOUS PURPOSES | 2022 |
|
RU2794525C1 |
DETECTING AND IDENTIFYING OBJECTS ON IMAGES | 2020 |
|
RU2726185C1 |
TEXTURED NEURAL AVATARS | 2019 |
|
RU2713695C1 |
METHOD FOR PROCESSING COMPUTER TOMOGRAPHY IMAGES (CT IMAGES) | 2023 |
|
RU2812866C1 |
METHOD OF AUTOMATIC SEGMENTATION OF FLUOROGRAMM OF BREAST OF PATIENTS WITH PNEUMONIA | 2016 |
|
RU2629629C1 |
APPARATUS AND METHOD FOR IMAGE PROCESSING AND SYSTEM FOR TRAINING A NEURAL NETWORK | 2019 |
|
RU2762144C1 |
DEVICE FOR COLOURING BLACK AND WHITE IMAGE | 2013 |
|
RU2524869C1 |
METHOD FOR DETERMINATION OF TOTAL CLOUD SCORE BY COLOR DIGITAL WIDE-ANGLE IMAGES OF VISIBLE SKY HEMISPHERE BASED ON STATISTICAL DATA PROCESSING METHODS | 2020 |
|
RU2777493C2 |
HARDWARE-SOFTWARE COMPLEX DESIGNED FOR TRAINING AND (OR) RE-TRAINING OF PROCESSING ALGORITHMS FOR AERIAL PHOTOGRAPHS OF THE TERRITORY FOR DETECTION, LOCALIZATION AND CLASSIFICATION UP TO TYPE OF AVIATION AND GROUND EQUIPMENT | 2020 |
|
RU2747044C1 |
Authors
Dates
2025-05-21—Published
2024-05-28—Filed