Sulama/İlaçlama Robotu için Nesne Tanıma Çalışmaları

Öz

Yapay zekâ ve buna bağlı alt çalışma konularındaki bilimsel yöntemler, neredeyse her alanda yaygın biçimde kullanılmaktadır. Özellikle savunma sanayi ve tarım bu alanların en cazip uygulama merkezlerini oluşturmaktadırlar. Tarımda veya bahçelerde kullanılmak üzere daha önce geliştirmiş olduğumuz sulama ve ilaçlama prototip mobil robotumuzun, nesne (ağaç veya değil) tanıma kısmında yeterli düzeyde başarılı neticeler alınamamıştı. Bu çalışmada, bu mobil robot ile gerçek zamanlı görüntüler üzerinde karmaşık (asimetrik) yapıya sahip bitki nesnelerinin tanınması ve bu tanınma sonucuna göre ilaçlama mekanizmasının tetiklenmesinin başarımını artırılmıştır. Bitki nesnelerinin tanınması amacı ile Viola-Jones nesne tanıma algoritması kullanılmıştır. Bu algoritma ile yapılan eğitim ve parametrik düzenlemeler sonucunda elde edilen modelin 5 farklı senaryo yolu üzerinde sınanması sonucunda, mobil sulama/ilaçlama robotunun görme yetisi ve bitki nesnelerini tanıma başarısının yaklaşık olarak %96.5 gibi oldukça yüksek bir düzeye çıkarıldığı görülmüştür. Ayrıca, ihtiyaç dolayısı ile uygulanan, herhangi bir nesneye ait görüntü eğitim setinin oluşturulmasına yönelik hızlı ve kolay bir yöntemin detayları makalede açıklanmıştır.

Anahtar Kelimeler

• Mobil robotlar
• nesne tanıma
• görüntü seti oluşturma
• yapay zeka
• akıllı sistemler

Object Recognition Studies for Irrigation/Medication Robot

Öz

Scientific methods in artificial intelligence and related sub-studies are widely used in almost every field. Defense industry and agriculture are the most attractive application areas of these fields. Sufficient successful results could not be obtained in the object (tree or not) recognition part of our irrigation and spraying prototype mobile robot, which we developed earlier for use in agriculture or gardens. In this study, with this mobile robot, the performance of the identification of complex (asymmetric) plant objects on real-time images and the triggering of the spraying mechanism according to this recognition result has been increased. Viola-Jones object capture algorithm was used to identify plant objects. As a result of the testing the model obtained through the training and parametric arrangements made to this algorithm on 5 different scenario paths, it has been observed that the success of recognizing plant objects and vision ability of the mobile irrigation/medication robot has been increased to a very high level of approximately 96,5%. In addition, the details of a quick and easy method for creating an image training set for any object that is applied due to need are explained in the article.

Anahtar Kelimeler

• Mobile robots
• object detection
• creating image set
• artifical intelligence
• smart systems

Kaynakça

1. Liu Yun, Zhang Peng, An Automatic Hand Gesture Recognition System Based on Viola-Jones Method and SVMs, 2009 Second International Workshop on Computer Science and Engineering, pp. 72-75.
2. Ramana Isukapalli, Ahmed Elgammal, Russell Greiner (2006): Learning to detect objects of many classes using binary classifiers. 9th European Conference on Computer Vision (ECCV 2006). Vol. 3951, no.1, pp. 352-364.
3. Mohammad Mahdi M., Maryam N., Majid P. and Mohammad Kazem M. (2018): Moving Vehicle Detection Using AdaBoost and Haar-Like Feature in Surveillance Videos. International Journal of Imaging and Robotics. Vol. 18, no. 1.
4. Moad. Benkiniouar, Mohamed Benmohammed (2010): Optimisation the real time implementation of the Viola & Jones face detection algorithm on RIse processor. 2010 XIth International Workshop on Symbolic and Numerical Methods, Modeling and Applications to Circuit Design (SM2ACD). IEEE.
5. Yasaman Heydarzadeh, Abolfazl Toroghi Haghighat, Nazila Fazeli (2010): Utilizing Skin Mask and Face Organs Detection for Improving the Viola Face Detection Method. 2010 Fourth UKSim European Symposium on Computer Modeling and Simulation. Pp. 174-178, IEEE.
6. Rapee Krerngkamjornkit, Milan Simic (2013): Enhancement of human body detection and tracking algorithm based on Viola and Jones framework. 2013 11th International Conference on Telecommunications in Modern Satellite, Cable and Broadcasting Services (TELSIKS). Vol. 1, pp. 115-118, IEEE.
7. Dewiani Djamaluddin, Tantri Indrabulan, Andani, Indrabayu, Sitti Wetenriajeng Sidehabi (2014): The Simulation of Vehicle Counting System for Traffıc Surveillance Using Viola Jones Method. 2014 Makassar International Conference on Electrical Engineering and Informatics (MICEEI). pp. 130-135, IEEE.
8. Yongzheng Xu, Guizhen Yu, Member, IEEE, Xinkai Wu, Yunpeng Wang, and Yalong Ma (2017): An Enhanced Viola–Jones Vehicle Detection Method From Unmanned Aerial Vehicles Imagery. IEEE Transactions on Intelligent Transportation Systems. Vol. 18, no. 7, pp. 1845-1856. IEEE.

9. Muhammad R Wiratama, Sukmawati N Endah, Retno Kusumaningrum, Helmie A Wibawa (2017): Pornography Object Detection Using Viola-Jones Algorithm and Skin Detection. 2017 1st International Conference on Informatics and Computational Sciences (ICICoS). P. 29-34. IEEE.
10. Kartika Candra Kirana, Slamet Wibawanto, Heru Wahyu Herwanto (2018): Facial Emotion Recognition Based on Viola-Jones Algorithm in the Learning Environment. 2018 International Seminar on Application for Technology of Information and Communication (iSemantic). Pp. 406-410. IEEE.
11. Pooya Tavallali, Mehran Yazdi, Mohammad Reza Khosravi (2017): An Efficient Training Procedure for Viola-Jones Face Detector. 2017 International Conference on Computational Science and Computational Intelligence (CSCI). Pp. 828-831. IEEE.
12. Benny Hardjono, Hendra Tjahyadi, Mario G. A. Rhizma, Andree E. Widjaja, Roberto Kondorura, Andrew M. Halim (2018): Vehicle Counting Quantitative Comparison Using Background Subtraction, Viola Jones and Deep Learning Methods. 2018 IEEE 9th Annual Information Technology, Electronics and Mobile Communication Conference (IEMCON). Pp. 556-562. IEEE.
13. LU Wen-yaoYang Ming (2019): Face Detection Based on Viola-Jones Algorithm Applying Composite Features. 2019 International Conference on Robots & Intelligent System (ICRIS). Pp. 82-85. IEEE.
14. Paul Viola Michael Jones (2001): Rapid Object Detection using a Boosted Cascade of Simple Features. Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001. Pp. 511-518. IEEE.
15. Haluk Özgen, Metin Turan (2018): Camera Assisted Mobile Medication and Irrigation Robot Design. ELECO 2018, Elektrik-Elektronik ve Biyomedikal Mühendisliği Konferansı. Pp. 162-168.