To solve the curse of dimensionality problem in multi-agent reinforcement learning, a learning method based on k-means is presented in this paper. In this method, the environmental state is represented as key state factors. The state space explosion is avoided by classifying states into different clusters using k-means. The learning rate is improved by assigning different states to existent clusters, as well as corresponding strategy. Compared to traditional Q-learning, our experimental results of the multi-robot cooperation show that our scheme improves the team learning ability efficiently. Meanwhile, the cooperation efficiency can be enhanced successfully.
In this paper, a hybrid control system is incorporated into a single agent structure. Two important features of our model include that the behaviors of terrain mapping are a set of hybrid primitives inside each agent, and the coordination strategies are implemented hierarchically in multiagent level. Thus, the key contribution of our work is that the model handles the high level requests as well as low level requests consistently in multi-robot terrain mapping. Moreover, it greatly facilitates mathematical analysis. Therefore, the analysis of stability is formulated for the terrain mapping controls, which guarantees the implemented systems stable within certain boundaries. The simulation of our system is implemented using Deterministic Finite Automaton in order to be consistent with the hybrid dynamics. The implemented system verifies the feasibility and efficiency of our model.
We proposed an online method of tracking the tunnel cable based on egomotion estimation. The method is firstly applied key point detection algorithm to extract feature points, and then the points are matched to estimate the matrix of egomotion representing the camera movement. Finally, we use the matrix to locate a mask around the cable in each frames captured inside the power line tunnel. The experimental results show robustness and efficiency of our method.
The traditional background difference method is based on gray image. Some information is lost when color image is transformed into gray image. So it is difficult to discriminate different colors with similar gray values and easily disturbed by noise and shadows. In this paper, the background difference is based on RGB color model. It is proposed to use the average value of each pixel of the color image sequences to extract the background, and then use the three-dimensional color values of the current frame and background image to compute the difference to detect the moving objects. The proposed approach is simple and easy to implement. The experimental results show that it is more sensitive to colors and has higher accuracy and robustness than the traditional background difference method. Besides, it is more resistant to shadows.
The process of tracking objects in real time video is one of the important topics in the study of suveillance system (Dhananjaya, Rama, and Thimmaiah 2015). Detection and extraction of information and tracking of objects or moving objects is as one form of application from computer vision. Some applications that use the method of tracking objects or moving objects include UAV (Unmanned Aerial Vehicle) surveillance or better known as the engine / vehicle unmanned, Indoor Monitoring system is an indoor monitoring system, as well as monitoring traffic traffic that can observe the movement of all Objects in real time state. Tracking objects in real time state many things to note and need to be taken into account where all the parameters and noise or disturbance of the surrounding objects that we do not need to observe but still in one part with the object that we observe. In this research the method to be used is background subtraction For the detection of tracking object And moving objects in real time based on color By utilizing webcam cameras and using OpenCv opensource library. 1. Pendahuluan Dalam perkembangan teknologi informasi dan komunikasi modern muncul bidang yang mempelajari tentang computer vision. Computer vision merupakan salah satu bagian dari bidang teknologi informasi dan komunikasi yang merupakan perkembangan dari ilmu grafika serta pengolahan citra digital yang bergabung dengan berbagai ilmu bidang komputer yang di tujukan untuk meniru cara kerja dari pengelihatan manusia yang dapat menangkap berbagai informasi diantaranya geometri, warna, ukuran, warna dan interpretasi dari suatu obyek. Kajian di bidang Computer vision telah berkembang pesat di barbagai bidang diantaranya bagian militer, kesehatan, industri dan lain-lain degan banyak ditemukanya peralatan yang canggih yang di hasilkan dari kajian computer vision antara lain yang menyangkut biomatric seperti deteksi wajah orang, sidik jari, retina mata serta yang non biomatric seperti pengenalan barang dengan sinar x, deteksi plat nomor kendaraan, jenis dan ukuran kendaraaan serta banyak lagi hal lain yang ditemukan dari kajian mengenai Computer Vision (Szeliski 2010). Deteksi objek berdasarkan warna merupakan suatu kajian yang sangat menarik yang dapat di implementasikan kedalam berbagai kehidupan baik dunia industi misalnya dalam pendeteksian kematangan dan pensortiran buah-buahan berdasarkan warna, quality control secara otomatis pada industri kain, kertas dan sebagainya yang memerlukan pendeteksian berdasarkan warna adalah sangat banyak, meskipun tidak menutup kemungkinan dikembankan pada industri sensorik
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.