本次研究為了能夠讓廠內檢查人員及工程師能夠快速辨識產品品質狀況，此次研究將透過Matlab影像處理強大的功能，將收集到的照片去做數位影像處理包含下列的功能(二值化、形態學、邊緣處理、熵值的分析等)，利用這些功能將AMB基板的缺陷辨識可以更為快速，加快產線檢查的時間，也可以將分析出的照片給工程師做缺陷的改善做為一個參考的指標。作為後續努力的目標，將會利用此次研究結果，也進而發展類神經網路或者其他小波處理進行影像辨識的學習及辨識的驗證。;With the advent of Global Industry 4.0, the use of big data, the Internet of Things, and robots is also in full swing. How to improve the production capacity and quality of the factory is already the most important issue for every company. The company must transform or be submerged in the torrent of times.
In this study, in order to enable the inspectors and engineers in the factory to quickly identify the quality status of the product, we used Matlab image processing powerful functions to process the collected photos for digital image processing, including the following functions (binarization, morphology, edge processing, analysis of entropy value, etc.), these functions can be used to identify the defects of the AMB heat dissipation substrate faster and speed up the inspection time of the production line for index. The next goal of our efforts will be to use the results of this research, and then to develop neural network or other wavelet processing for image recognition learning and recognition verification.
