Display Screen Quality Detection Using Deep Convolutional Neural Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing display screen quality detection methods are heavily influenced by subjective human factors, resulting in low accuracy, poor system performance, and limited business expansion capabilities.
Innovation Solution
A display screen quality detection method utilizing a deep convolutional neural network structure and instance segmentation algorithm for defect detection, which preprocesses images and determines quality based on historical defect data, reducing human influence and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual detection method is used, then system complexity is low, but detection accuracy is low and productivity is low
Solution Approach 1:
The patent replaces manual mechanical detection with an automated deep learning-based detection system. The defect detection model uses convolutional neural networks to automatically identify and classify defects, substituting human visual inspection with intelligent algorithmic analysis, thereby improving detection accuracy while maintaining manageable system complexity through modular architecture.
Solution Approach 2:
The patent introduces an intermediary defect detection model that acts as a bridge between raw display screen images and final quality decisions. This model processes images through multiple layers of feature extraction and defect classification, serving as an intelligent mediator that enhances detection accuracy without requiring direct human intervention in every inspection step.
2Measurement precision
If machine-assisted manual detection method is used, then productivity is improved, but detection accuracy is still affected by subjective factors
Solution Approach 1:
The patent implements a self-service detection system where the defect detection model autonomously performs defect identification, classification, and quality assessment without requiring human experts to manually examine each image. The system automatically processes display screen images, detects defects, and generates quality results, eliminating subjective human factors while maintaining high productivity through automated batch processing.
Solution Approach 2:
The patent transforms the detection process by changing key parameters from subjective human judgment to objective algorithmic metrics. The defect detection model uses standardized defect categories, automated classification thresholds, and consistent evaluation criteria, converting variable human perception into stable, repeatable measurement parameters that improve both accuracy and system performance.
3Measurement precision
If deep convolutional neural network and instance segmentation algorithm are used, then detection accuracy is high, but device complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the defect detection task into distinct functional modules: image preprocessing, defect detection, defect classification, and quality determination. The instance segmentation algorithm specifically segments different defect instances within images, allowing independent analysis and classification of each defect type, thereby managing model complexity through functional decomposition.
Solution Approach 2:
The patent creates a universal defect detection model that handles multiple defect types and categories through a single integrated system. The model is designed to detect various defects (dead pixels, line defects, surface defects, etc.) and classify them into standardized categories, providing multi-functional capability that reduces overall system complexity compared to having separate specialized detectors for each defect type.
Data Source
AI summary
A display screen quality detection method, an apparatus, an electronic device and a storage medium. The method includes receiving a quality detection request sent by a console deployed on a display screen production line, where the quality detection request includes a display screen image captured by an image capturing device on the display screen production line, performing image preprocessing on the display screen image, and inputting the preprocessed display screen image into a defect detection model to obtain a defect detection result, where the defect detection model is obtained by training with a historical defect display screen image using a deep convolutional neural network structure and an instance segmentation algorithm, determining, according to the defect detection result, quality of a display screen corresponding to the display screen image. The technical solution has high defect detection accuracy, good system performance, and high business expansion capability.