Search

CN-121980442-A - Die casting machining intelligent detection system and method integrating visual recognition

CN121980442ACN 121980442 ACN121980442 ACN 121980442ACN-121980442-A

Abstract

The application provides a die casting processing intelligent detection system and method integrating visual recognition, which are used for carrying out combined pretreatment on surface visual image data of nodes in different working procedures and physical signals of die casting materials in the processing process of a target die casting to obtain surface effective data and physical effective data related to die casting defects; extracting surface defect characteristics of a target die casting in the surface effective data, extracting physical defect characteristics of a target die casting material in the physical effective data, fusing the surface defect characteristics and the physical defect characteristics into associated fusion defect characteristics of the target die casting, acquiring space-time coordinate information of a sensor network, and determining defect characterization information comprising defect types and defect positions of processing defects of the target die casting according to the associated fusion defect characteristics and the space-time coordinate information. By adopting the scheme of the application, the defect detection can be carried out on the processing of the die casting based on the cooperative correlation between the surface and the physical defect characteristics.

Inventors

  • SUN BIN

Assignees

  • 盐城东创精密制造有限公司

Dates

Publication Date
20260505
Application Date
20251219

Claims (10)

  1. 1. An intelligent detection method for die casting processing integrating visual recognition, wherein, deploying a sensor network comprising a visual sensor and a physical parameter sensor on a target die casting accessory in advance, wherein the method comprises the following steps: acquiring surface visual image data of nodes in different procedures and die casting material physical signals in the processing process of a target die casting through a sensor network; performing joint pretreatment on the surface visual image data and the die casting material physical signals based on the historical defect characteristics corresponding to the target die casting to obtain surface effective data and physical effective data related to die casting defects; Extracting surface defect characteristics of a target die casting in the surface effective data, extracting physical defect characteristics of a target die casting material in the physical effective data, and carrying out association fusion on the surface defect characteristics and the physical defect characteristics to obtain association fusion defect characteristics of the target die casting; And acquiring space-time coordinate information of the sensor network, and determining defect characterization information comprising defect types and defect positions of the target die casting processing defects according to the associated fusion defect characteristics and the space-time coordinate information.
  2. 2. The method of claim 1, wherein the die cast material physical signal comprises a die cast process temperature, internal stress, material density, surface hardness.
  3. 3. The method of claim 1, wherein performing joint preprocessing on the surface visual image data and the die casting material physical signal based on the historical defect characteristics corresponding to the target die casting to obtain surface effective data and physical effective data related to the die casting defect specifically comprises: Acquiring historical defect characteristics corresponding to a target die casting; performing exception processing on the surface visual image data and the die casting material physical signal to obtain exception processed surface visual image data and die casting material physical signal; And carrying out cross verification on the surface visual image data subjected to the exception processing and the die casting material physical signal based on the association rule between the surface defect and the physical defect in the historical defect characteristic, so as to obtain surface effective data and physical effective data related to the die casting defect.
  4. 4. The method of claim 1, wherein extracting the surface defect features of the target die casting from the surface valid data specifically comprises: extracting a suspected defect area of the target die casting from the surface effective data; Determining geometric features, texture features and contour features of the target die casting according to the suspected defect region; And screening out the surface defect characteristics of the target die casting from the geometric characteristics, the texture characteristics and the contour characteristics based on the historical defect characteristics corresponding to the target die casting.
  5. 5. The method of claim 1, wherein extracting physical defect characteristics of the target die casting material in the physical effective data specifically comprises: extracting data characteristics of each dimension from the physical effective data; and extracting physical defect characteristics of the material of the target die casting from all the data characteristics based on the historical defect characteristics corresponding to the target die casting.
  6. 6. The method of claim 1, wherein performing an associative fusion of the surface defect feature and the physical defect feature to obtain an associative fusion defect feature of the target die casting specifically comprises: Performing time sequence alignment on the surface defect characteristics and the physical defect characteristics to construct a characteristic time sequence correlation matrix; Acquiring a correlation rule between the surface defect and the physical defect in the history defect characteristic; and carrying out layered fusion on the characteristic time sequence incidence matrix based on the incidence rule to obtain the incidence fusion defect characteristic of the target die casting.
  7. 7. The method of claim 1, wherein determining defect characterization information including a defect type and a defect location of a target die casting machining defect based on the associated fused defect signature and the space-time coordinate information comprises: pre-constructing a defect type-association fusion feature mapping library; performing similarity comparison on the associated fusion defect characteristics and the defect type-associated fusion characteristic mapping library to obtain a defect type of the target die casting machining defect; Positioning actual coordinate information of the surface defect of the target die casting according to the space-time coordinate information conversion and the associated fusion defect characteristic; And determining defect characterization information including the defect type and the defect position of the target die casting processing defect according to the defect type and the actual coordinate information.
  8. 8. Die casting processing intelligent detection system that fuses visual identification, a serial communication port, include: the acquisition module is used for acquiring surface visual image data of nodes in different procedures and physical signals of die casting materials in the processing process of the target die casting through the sensor network; The processing module is used for carrying out joint pretreatment on the surface visual image data and the die casting material physical signals based on the historical defect characteristics corresponding to the target die casting to obtain surface effective data and physical effective data related to the die casting defects; the processing module is further used for extracting surface defect characteristics of the target die casting in the surface effective data, extracting physical defect characteristics of the material of the target die casting in the physical effective data, and carrying out association fusion on the surface defect characteristics and the physical defect characteristics to obtain association fusion defect characteristics of the target die casting; and the execution module is used for acquiring the space-time coordinate information of the sensor network, and determining defect characterization information comprising the defect type and the defect position of the processing defect of the target die casting according to the associated fusion defect characteristic and the space-time coordinate information.
  9. 9. A computer device comprising a memory storing code and a processor configured to obtain the code and perform the die casting machining intelligent detection method of fusion visual recognition according to any one of claims 1 to 7.
  10. 10. A computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the die casting machining intelligent detection method of fusion visual recognition according to any one of claims 1 to 7.

Description

Die casting machining intelligent detection system and method integrating visual recognition Technical Field The application relates to the technical field of processing detection, in particular to an intelligent detection system and method for processing a die casting by integrating visual recognition. Background The machining detection refers to the process of checking and measuring the characteristics of the size, shape, position, surface quality and the like of the machined parts by using precision measurement equipment and technical means, and is a core link for ensuring the product quality in a modern manufacturing system, and whether the machined parts accord with design drawings and process requirements is verified by detecting the machined parts in real time or afterwards. The die casting is widely applied to the high-end manufacturing fields of automobiles, aerospace and the like due to high molding efficiency and strong suitability of mechanical properties, the surface quality and the internal quality of the die casting directly determine the reliability of a terminal product, efficient and accurate processing detection becomes an industrial core requirement, the traditional detection mode has obvious limitations that manual visual detection depends on experience and is low in efficiency and easy to miss detection of fine defects, single visual sensing can only capture surface morphology anomalies and can not identify internal physical defects such as loose materials and concentrated stress, single physical parameter sensing (such as temperature and stress sensors) can monitor material characteristics, but is difficult to correlate the specific morphology and position of the surface defects, meanwhile, the prior art does not fully utilize historical defect data to guide detection, the effective data extraction rate in a pretreatment stage is low, the surface and physical defect characteristics are isolated and the cooperative correlation information of the surface and the physical defect characteristics is lost, and therefore how to detect the defects of processing of the die casting based on the cooperative correlation between the surface and the physical defect characteristics becomes an industrial problem. Disclosure of Invention The application provides an intelligent detection system and method for processing a die casting by integrating visual recognition, which can detect defects in the processing of the die casting based on cooperative correlation between surface and physical defect characteristics. In a first aspect, the present application provides an intelligent detection method for die casting processing with fusion of visual recognition, wherein a sensor network including a visual sensor and a physical parameter sensor is deployed in advance on a target die casting accessory, and the method comprises the following steps: acquiring surface visual image data of nodes in different procedures and die casting material physical signals in the processing process of a target die casting through a sensor network; performing joint pretreatment on the surface visual image data and the die casting material physical signals based on the historical defect characteristics corresponding to the target die casting to obtain surface effective data and physical effective data related to die casting defects; Extracting surface defect characteristics of a target die casting in the surface effective data, extracting physical defect characteristics of a target die casting material in the physical effective data, and carrying out association fusion on the surface defect characteristics and the physical defect characteristics to obtain association fusion defect characteristics of the target die casting; And acquiring space-time coordinate information of the sensor network, and determining defect characterization information comprising defect types and defect positions of the target die casting processing defects according to the associated fusion defect characteristics and the space-time coordinate information. In some embodiments, the die cast material physical signal includes a die cast process temperature, an internal stress, a material density, a surface hardness. In some embodiments, performing joint preprocessing on the surface visual image data and the die casting material physical signal based on the historical defect characteristics corresponding to the target die casting, and obtaining the surface valid data and the physical valid data related to the die casting defect specifically includes: Acquiring historical defect characteristics corresponding to a target die casting; performing exception processing on the surface visual image data and the die casting material physical signal to obtain exception processed surface visual image data and die casting material physical signal; And carrying out cross verification on the surface visual image data subjected to the exception processing and the die casti