Real-Time Multisensor Vehicle Localization: A Geographical Information System?Based Approach

Aynaud, Claude; Bernay-Angeletti, Coralie; Aufrère, Romuald; Lequièvre, Laurent; Debain, Christophe; Chapuis, Roland

doi:10.1109/mra.2017.2669399

Cited by 5 publications

(1 citation statement)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Object tracking is developed to recognize objects and estimate their position at each time change. The existence of object tracking is closely related to other fields, such as distance learning [1], [2], surveillance [3]- [8], robotic navigation [9], vehicle localization [10], [11], and smart cities [3]. One of the problems in object tracking algorithms include the need to ensure that they follow objects in real time and that their resistance to changes in environmental conditions, such as lighting, is increased.…”

Section: Introductionmentioning

confidence: 99%

Mean-shift Object Tracking Algorithm with Systematic Sampling Technique

Bandung¹,

Ardiansyah²

2021

MST

View full text Add to dashboard Cite

Mean shift is a fast object tracking algorithm that only considers pixels in an object area, hence its relatively small computational load. This algorithm is suitable for use in real-time conditions in terms of execution time. The use of histograms causes this algorithm to be relatively resistant to rotation and changes in object size. However, its resistance to lighting changes is not optimal. This study aims to improve the performance of the algorithm under lighting changes and reduce its processing time. The proposed technique involves the use of sampling techniques to reduce the number of iterations, optimization of candidate search object locations using simulated annealing, and addition of tolerance parameter to optimize object location search and area-based weighting instead of the Epanechnikov kernel. The results of the one-tail t-test with two independent sample groups reveal that the average performance of the proposed algorithm is significantly better than that of the traditional mean-shift algorithm in terms of resistance to lighting changes and processing time per video frame. In the test involving 999 frames of video images, the average processing time of the proposed algorithm is 83.66 ms, whereas that of the traditional mean-shift algorithm is 116.86 ms. AbstrakAlgoritma Pelacakan Objek Berbasis Mean-Shift Dengan Sistematik Sampling. Mean-shift adalah algoritma pelacakan objek yang cepat karena hanya mempertimbangkan piksel pada area objek. Selain itu, algoritma ini memiliki beban komputasi yang lebih kecil sehingga sangat sesuai untuk digunakan dalam kondisi waktu nyata dalam hal waktu eksekusi. Penggunaan histogram citra menyebabkan algoritma ini relatif tahan terhadap rotasi dan perubahan ukuran objek. Namun, daya tahan algoritma terhadap perubahan pencahayaan belum optimal. Penelitian ini bertujuan untuk memperkenalkan penggunaan teknik sampling pixel untuk mengurangi beban data yang perlu diolah dalam pemrosesan dan meningkatkan kinerja algoritma dalam kondisi perubahan pencahayaan. Teknik yang diusulkan meliputi penggunaan teknik pengambilan sampel untuk mengurangi jumlah iterasi, optimalisasi lokasi objek pencarian kandidat menggunakan Algoritma Simulated Annealing, penambahan parameter toleransi untuk mengoptimalkan pencarian lokasi objek, dan pembobotan berbasis area sebagai pengganti kernel Epanechnikov. Dengan menggunakan statistik uji one tail t-test dengan dua kelompok sampel independen, hasil tes menunjukkan bahwa rata-rata kinerja algoritma yang diusulkan lebih cepat secara signifikan daripada algoritma mean-shift dalam hal waktu pemrosesan per frame video dan ketahanan terhadap perubahan pencahayaan. Hasil pengujian dengan 999 frame gambar video memberikan rata-rata waktu pemrosesan dari algoritma yang diusulkan adalah 83,66 ms sedangkan algoritma mean-shift adalah 116,86 ms.

show abstract