Neighboring Predictive Gradient Spatio-Temporal Sequencing Algorithm: An Object Sequencing Algorithm for Logical Ordering of Sparse Object Detections

When object detection is carried out in settings with sparse and irregular data acquisition, conventional sequencing techniques that depend on continuous tracking or dense observations frequently fall short of reconstructing the proper logical sequence of events. This paper introduces the Neighborin...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE access Jg. 13; S. 160034 - 160047
Hauptverfasser: Yeap, Herrick Han Lin, Eu, Kok Seng, Tan, Tee Hean, Yap, Kian Meng
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Piscataway IEEE 2025
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:
Accuracy
Algorithms
clustering
Clustering algorithms
Clustering methods
Data acquisition
Data points
data sparsity
Encoding
first-in-first-out (FIFO)
moving object detection
object classification
Object detection
Object recognition
Prediction algorithms
Sensors
Sequence detection
Sequences
Sequential analysis
Sparsity
Trajectory
ISSN:2169-3536, 2169-3536
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:When object detection is carried out in settings with sparse and irregular data acquisition, conventional sequencing techniques that depend on continuous tracking or dense observations frequently fall short of reconstructing the proper logical sequence of events. This paper introduces the Neighboring Predictive Gradient Spatio-Temporal Sequencing Algorithm (NPGSTSA), a novel framework for determining the sequential order of detected objects in sparse and scarce environments. NPGSTSA leverages the relative y-intercepts of detected objects and their neighboring relations as proxies for sequence position inference. By utilizing the properties of the gradient and its neighboring data object, the algorithm is capable of robustly estimating sequence flow even under severe data sparsity and insufficient data points. To evaluate its effectiveness, we constructed a simulated video-based object detection with varying data sparsity levels and downsampling factors. To assess the quantitative accuracy of the sequence, we modified a 1-Wasserstein distance as a measurement metric, demonstrating that NPGSTSA significantly outperforms conventional methods, such as First-In-First-Out (FIFO) and cluster-based sequencing. The results confirm the algorithm's capacity to infer coherent object sequences even in data-constrained scenarios, highlighting its practical applicability, such as logistics environments where operational speed and efficiency are critical.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2169-3536
2169-3536
DOI:10.1109/ACCESS.2025.3604022