Abstract

In this study, an algorithm was developed that can detect and present number of perforated and non-perforated objects on an image directly or on an image which was captured momentarily by a camera. Firstly, the algorithm was developed using C programming language on Matlab environment and its performance was tested on the artificially produced binary images. At the same time, a developed version of the algorithm was created on Visual Studio Code environment using Python programming language that is used frequently in many software. After successful trials on the prepared artificial images, a real-time object detection algorithm was achieved with inputs of algorithm provided with images taken momentarily by camera. To use the developed algorithm at different applications, Raspberry-Pi based experimental setup was prepared from single board computer platforms and the developed algorithm was run on. In conclusion, it was seen that the number and the exact positions of perforated and non-perforated objects on the images can be detected successfully.

References

[1] K. Yaman, N. Aktürk. Görüntü işleme ile kişi yoğunluklarının belirlenmesi. 10. Ulusal Makine Teorisi Sempozyumu 2001.

[2] İ. Soyhan, S. Gürel, S.A. Tekin. Yapay zeka tabanlı görüntü işleme tekniklerinin insansız hava araçları üzerinde uygulamaları. European Journal of Science and Technology Special Issue 2021; 24:473-469.

[3] A. Eldem, H.Eldem, A. Palalı. Görüntü işleme teknikleriyle yüz algılama sistemi geliştirme. BEU Journal of Science 2017; 6(2):48-44. 

[4] A.E. Aytan, Y. Öztürk, E.K. Örgev. Görüntü işleme. İ.Ü. Diş Hekimliği Fakültesi Dergisi 1993; 27(4):277-273

[5] K. Hanbay, H. Üzen. Nesne tespit ve takip metotları: Kapsamlı bir derleme. Turkish Journal of Nature and Science 2017; 6(2):49-40.

[6] L.A. Szolga. On flight real time image processing by drone equipped with raspberry pi4. International Symposium for Design and Technology in Electronic Packaging 2021; p. 334-337.

[7] M. Ariyanto, I. Haryanto, J.D. Setiawan, M. Munadi, M.S. Radityo. Real-time image processing method using raspberry pi for a car model. International Conference on Electric Vehicular Technology 2019; p. 46-51.