Unlock instant, AI-driven research and patent intelligence for your innovation.

Assembling connecting piece image feature deep learning and identification method

A deep learning and image feature technology, applied in image enhancement, image analysis, image data processing and other directions, can solve problems such as limited use range, poor flexibility, and inability to detect

Active Publication Date: 2017-08-04
GUANGZHOU HUAJIE ELECTRONICS TECH CO LTD
View PDF7 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The recognition method based on inherent features is good for the detection of specific parts. If the object changes, the features must be re-detected by professional visual inspection personnel, which is less flexible.
The detection method based on matching can be more flexible, and it cannot detect the situation that has not been set in advance, which limits its scope of application

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Assembling connecting piece image feature deep learning and identification method
  • Assembling connecting piece image feature deep learning and identification method
  • Assembling connecting piece image feature deep learning and identification method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0014] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be described in further detail below with reference to the embodiments and accompanying drawings.

[0015] like figure 1 As shown, it is a process of deep learning and recognition method for assembling connector image features, and the method includes the following steps:

[0016] Step 1 sample collection stage, collect qualified image sequence (I 1 ,I 2 …I n ), the test sample range s specified by the test personnel c , sample type C(s c ), calculate the maximum displacement in the horizontal direction Maximum displacement with vertical direction Construct the co-training matrix M X , M Y , get the co-sample set S c .

[0017] Step 2 In the sample deep learning stage, the co-sample set S is c Input the convolutional neural network for training, and calculate the bias matrix P of the co-sample set c , modify the co-training matri...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an assembling connecting piece image feature deep learning and identification method. The method includes a sample acquisition stage, a sample deep learning stage, and an identification stage. The sample acquisition stage includes acquiring the qualified image, allowing the test personnel to specify the test sample range, the sample type, the calculation level, and the vertical maximum displacement, constructing an associated training matrix, and obtaining an associated sample set; the sample deep learning stage includes inputting the associated sample set into a convolution neural network for training, calculating the deviation matrix of the associated sample set, modifying the associated training matrix and performing multitime iterations to make the deviation matrix meet the threshold requirement; and the identification phase includes inputting the sample to be identified into the convolution neural network obtained in the deep learning stage and outputting the class. The method can realize the automatic identification of the various connecting pieces of the assembly process, and can realize the identification after the automatic learning by artificially specifying the sample position and type, thereby realizing the intelligent and integrated assembling quality detection for different workpieces.

Description

technical field [0001] The invention relates to a method for recognizing an assembly connector, in particular to a method for deep learning and recognition of image features of an assembly connector. Background technique [0002] Assembly refers to the process of fitting and connecting parts or components to make them semi-finished or finished products according to the specified technical requirements. Assembly is an important process in the product manufacturing process. The quality of the assembly plays a decisive role in the quality of the product. The discussion of assembly quality is carried out on the premise that all the assembled parts are qualified and installed correctly. Therefore, before testing the assembly quality of the product, it is necessary to check whether there are wrong or missing parts. Due to the large number of assembly parts and connectors, and the diversification of assembly inspection objects, higher requirements and challenges are raised for the...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06K9/46G06K9/62G06T7/00G06T7/13
CPCG06T7/0004G06T2207/30164G06V10/44G06F18/214
Inventor 林镇秋黄瑛娜杨锦波
Owner GUANGZHOU HUAJIE ELECTRONICS TECH CO LTD