Anchor-frame-free target detection network for electronic component and detection method applying network

Abstract

The invention discloses an anchor-frame-free target detection network for an electronic component and a detection method applying the network. The anchor-frame-free target detection network for the electronic component comprises an Encoder part, a Decoder part and a Head part, wherein the Encoder part comprises an improved backbone feature extraction network EfficientNet, the EfficientNet comprises an MBConv-ECA module and a variable convolution module, the Decoder part comprises a deconvolution module for realizing an up-sampling function, the Head part comprises a classification prediction and bounding box regression prediction module, and after classification prediction and bounding box regression prediction, loss calculation and back propagation are performed on a forward propagation result and a true value of the anchor-frame-free target detection network, so that a gradient is updated, and network training is completed. Anchor-frame-free target detection of the electronic component is achieved through using the network, the category and position information of the electronic component can be obtained in real time, and real-time detection of the electronic component is achieved.

Description

technical field [0001] The invention belongs to the field of identification and detection of electronic components, in particular to an anchor frameless target detection network for electronic components and a detection method using the network. Background technique [0002] The identification and detection of electronic components is an important part of realizing the intelligent assembly of electronic components. The identification and detection technology for electronic components has also developed rapidly. The identification and detection of electronic components in the prior art mainly focus on the following two methods: one is based on traditional template matching and feature point matching to realize the identification and positioning of electronic components, and the other is based on deep learning target detection algorithm Realize the identification and positioning of electronic components. [0003] However, the above recognition and positioning technologies st...

AI Technical Summary

Problems solved by technology

[0003] However, the above identification and positioning technologies still have the following defects: 1. Although the traditional template matching and feature point matching are fast, their accuracy is not high and their robustness is poor, and they have high requirements for lighting and image shooting angles.

2. The target detection algorithm based on deep learning realizes the detection of electronic components based on the anchor frame, and the size setting of the anchor frame will directly affect the final detection effect to a large extent

Images