A Robust Real-Time Detecting and Tracking Framework for Multiple Kinds of Unmarked Object.

Sensors (Basel)

Beijing Advanced Innovation Center for Intelligent Robot and System, Beijing Institute of Technology, Beijing 100081, China.

Published: December 2019

A rodent real-time tracking framework is proposed to automatically detect and track multi-objects in real time and output the coordinates of each object, which combines deep learning (YOLO v3: You Only Look Once, v3), the Kalman Filter, improved Hungarian algorithm, and the nine-point position correction algorithm. A model of a Rat-YOLO is trained in our experiment. The Kalman Filter model is established in an acceleration model to predict the position of the rat in the next frame. The predicted data is used to fill the losing position of rats if the Rat-YOLO doesn't work in the current frame, and to associate the ID between the last frame and current frame. The Hungarian assigned algorithm is used to show the relationship between the objects of the last frame and the objects of the current frame and match the ID of the objects. The nine-point position correction algorithm is presented to adjust the correctness of the Rat-YOLO result and the predicted results. As the training of deep learning needs more datasets than our experiment, and it is time-consuming to process manual marking, automatic software for generating labeled datasets is proposed under a fixed scene and the labeled datasets are manually verified in term of their correctness. Besides this, in an off-line experiment, a mask is presented to remove the highlight. In this experiment, we select the 500 frames of the data as the training datasets and label these images with the automatic label generating software. A video (of 2892 frames) is tested by the trained Rat model and the accuracy of detecting all the three rats is around 72.545%, however, the Rat-YOLO combining the Kalman Filter and nine-point position correction arithmetic improved the accuracy to 95.194%.

Download full-text PDF

Source
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC6982905PMC
http://dx.doi.org/10.3390/s20010002DOI Listing

Publication Analysis

Top Keywords

kalman filter
12
nine-point position
12
position correction
12
current frame
12
tracking framework
8
deep learning
8
correction algorithm
8
labeled datasets
8
frame
6
position
5

Similar Publications

Accurately estimating phase is crucial in continuous-variable quantum key distribution systems, directly impacting the final secret key rate. In previous systems that utilize the local local oscillator, phase estimation is closely tied to the amplitude and signal-to-noise ratio (SNR) of the pilot signal. As SNR decreases, so does the accuracy of phase estimation, leading to increased excess noise and a potential loss of the system's secret key rate.

View Article and Find Full Text PDF

State estimation for large-scale non-Gaussian dynamic systems remains an unresolved issue, given nonscalability of the existing particle filter algorithms. To address this issue, this paper extends the Langevinized ensemble Kalman filter (LEnKF) algorithm to non-Gaussian dynamic systems by introducing a latent Gaussian measurement variable to the dynamic system. The extended LEnKF algorithm can converge to the right filtering distribution as the number of stages become large, while inheriting the scalability of the LEnKF algorithm with respect to the sample size and state dimension.

View Article and Find Full Text PDF

To enhance the positioning accuracy of autonomous underwater vehicles (AUVs), a new adaptive filtering algorithm (RHAUKF) is proposed. The most widely used filtering algorithm is the traditional Unscented Kalman Filter or the Adaptive Robust UKF (ARUKF). Excessive noise interference may cause a decrease in filtering accuracy and is highly likely to result in divergence by means of the traditional Unscented Kalman Filter, resulting in an increase in uncertainty factors during submersible mission execution.

View Article and Find Full Text PDF

Topography estimation is essential for autonomous off-road navigation. Common methods rely on point cloud data from, e.g.

View Article and Find Full Text PDF

This paper presents a comparative study of different AI models for indoor positioning systems, emphasizing improvements in localization accuracy and processing time. This study examines Artificial Neural Networks (ANNs), Long Short-Term Memory (LSTM), Recurrent Neural Networks (RNNs), and the Kalman filter using a real Received Signal Strength Indicator (RSSI) and 9-axis ICM-20948 sensor. An in-depth analysis is provided in this paper for data cleaning and feature selection to reduce errors for all the models.

View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!