Battery Weld Inspection Using 2D-3D Fusion and AI Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies lack an efficient and accurate method for inspecting welding defects in secondary batteries, which are crucial for ensuring the quality and safety of batteries used in green technology and eco-friendly vehicles.
Innovation Solution
A welding inspection apparatus and method utilizing a scanner to acquire both two-dimensional and three-dimensional images, a data processor to generate fusion data, and an artificial intelligence model to identify welding areas and determine defects, incorporating preprocessing techniques like resizing, normalization, and weighted operations to enhance accuracy and robustness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional welding inspection methods are used, then the inspection process is simple, but the accuracy and reliability of welding defect detection is insufficient
Solution Approach 1:
The patent combines 2D imaging data and 3D imaging data into fused data, integrating multiple information sources to improve detection accuracy. The fusion of multi-dimensional data allows the system to overcome the limitations of single-modality inspection and achieve more reliable welding defect detection.
Solution Approach 2:
The patent introduces an artificial intelligence model as an intermediary between raw imaging data and defect determination. This AI model processes and interprets the fused 2D and 3D data, enabling accurate identification of welding areas and defects while managing the complexity of the inspection system.
2Measurement precision
If high-resolution images with more pixels are used, then the image quality and detail information are improved, but the data processing time and computational load increase
Solution Approach 1:
The patent applies normalization to selectively scale pixel value ranges (e.g., converting 16-bit height values to 8-bit normalized values). This partial transformation maintains the essential quality information needed for defect detection while reducing the computational burden and processing time.
Solution Approach 2:
The patent transforms image parameters through normalization operations, changing the bit depth and value ranges of pixel data. This parameter transformation preserves the critical quality information required for accurate welding inspection while optimizing the data for efficient processing by the AI model.
Data Source
AI summary
A welding inspection apparatus of the present disclosure includes a scanner configured to acquire a two-dimensional image and a three-dimensional image by photographing a battery, a data processor configured to generate fusion data based on the two-dimensional image and the three-dimensional image, an object identifier configured to identify a welding area in the fusion data based on an artificial intelligence model trained to identify an object, and a welding determiner configured to determine whether a weld joint of the battery is defective based on the welding area.