6 results about "Machine vision image processing" patented technology

This invention discloses an improved YOLOv8 strip steel burr detection model based on a fusion of a dynamic working condition perception module, belonging to the field of machine vision image processing technology. The specific steps are as follows: A feature extraction fusion network is constructed with CSPDarknet-53 as the backbone, combined with C2f and SPPF modules, employing multi-scale feature fusion and a decoupled detection head; a dynamic working condition perception module is constructed to collect real-time working condition parameters during strip steel production, which are then processed by Z-score standardization and dynamic weight allocation before being input into a dual-branch lightweight sub-network to generate adjustment signals for the detection threshold and anchor frame size; these adjustment signals are dynamically injected into the forward inference process of the YOLOv8 detection head through a hook function to achieve real-time adaptation of detection parameters to working conditions; the final model is obtained through multi-stage collaborative training. This invention significantly improves the robustness and generalization ability of burr detection under complex working conditions, achieving a detection accuracy of over 95% and a false negative rate of 3.1%, meeting the real-time detection needs of industry.

The invention discloses a method for processing a transparent melt based on a machine vision image and controlling crystal growth in real time. The method comprises the following steps: acquiring a video image of a crystal and a melt interface in a crystal growth process in real time through image acquisition equipment; preprocessing the acquired image and extracting effective features; based on a double-threshold segmentation image, determining an accurate identification range and identification accuracy, and in combination with a polar coordinate sector density analysis method, identifying the lower edge boundary of the crystal and calculating the radius of the crystal; recursive least square harmonic wave trapping is adopted to remove periodic interference signals caused by crystal rotation, and a balance diameter sequence reflecting the actual growth trend is obtained; according to the change of the balance diameter sequence along with time, calculating a diameter slope as a crystal growth rate prediction value; based on the growth rate and the crystal radius, a diameter-slope-power relation model is established, and real-time closed-loop regulation and control of the heating power are achieved; according to the invention, a function relationship is established between the crystal diameter change trend and the power correction, automatic correction of the temperature deviation is realized through a sine smooth power regulation function, and a closed-loop system of visual detection-trend analysis-power feedback is formed.

The invention relates to the technical field of machine vision image processing, and discloses a part surface defect detection method based on image processing. The method comprises the following steps: acquiring a part surface image through image acquisition, extracting defect sensitive characteristics by using a multi-scale Gaussian difference operator, and filtering through a threshold value to obtain a candidate defect point set; and inputting the candidate point set into a serialized network containing an attention mechanism and a memory unit to generate an enhanced feature sequence. And carrying out space alignment on the sequence and an original topological association graph, and generating an enhanced feature graph containing node attributes and edge weights through graph convolution. And on the basis, performing region growing and merging operation to generate a connected candidate defect region. And extracting geometric and texture descriptors of each region to form a feature vector, matching the feature vector with a standard vector in a defect feature library so as to identify the defect category and position, and finally outputting a defect detection map. According to the method, the detection rate and the segmentation precision of the complex distribution defects are improved.

This invention provides a method and system for reconstructing the outer contour function of a projectile's head shape based on machine vision, belonging to the field of machine vision image processing and automatic parameter calculation technology. First, the invention performs semantic segmentation on the frontal view image of the projectile's head shape to obtain a binary mask of the projectile target, and extracts and crops the outer contour of the head shape segment. Second, through coordinate axis self-calibration and attitude alignment, the standard attitude coordinates of the projectile's head shape outer contour are obtained. Then, in the local coordinate system, geometric constraint feature points are extracted to obtain the endpoints, pixel areas, and single-point coordinates of the projectile's head shape outer contour. Third, a two-parameter head shape outer contour shape function satisfying endpoint constraints is constructed, and the shape parameters are solved using area constraints and single-point constraints to obtain the analytical expression of the head shape outer contour. Finally, the outer contour function and its derivative are substituted into the head shape factor integral formula to achieve automatic calculation of the head shape factor.

A computer architecture for artificial intelligence model training allows for dual model machine-vision image processing that can reduce loss and computational load by acquiring training data including a training target image, a training reference image, and ground truth information for processing the training target image so that a training target pose of a training target object coincides with a training reference pose of a training reference object. A model calculates a first training target feature and a first training reference feature, and outputs first training processing information. A second model calculates a second training target feature and a second training reference feature, and outputs second training processing information based on the first training processing information, the second training target feature, and the second training reference feature.

The present application relates to the technical fields of precision manufacturing nondestructive testing and machine vision image processing, in particular to a titanium alloy ring surface micro-crack defect detection system based on machine vision, which comprises: an image acquisition module for obtaining image data sets to be processed; a manifold calibration module for obtaining the main direction field of background texture and transforming the image data sets to be processed into a standard space image aligned with the texture flow; a sparse decomposition module for obtaining shear wave coefficients and decomposing into a low-rank component matrix and a sparse component matrix; a reconstruction and discrimination module for generating a micro-crack defect distribution map, obtaining a residual image and generating a final defect distribution map; and an adaptive feedback module for adjusting the sparse constraint weight in the robust principal component analysis algorithm; the present application effectively overcomes the signal aliasing problem caused by the frequency overlap of cracks and background texture, and significantly improves the detection sensitivity under strong texture background.