Paper Info
Title
Thermal image processing-based intelligent technique for object detection
Abstract
The Thermal Image Processing Technique (TIPT) is the most prominent tool for observing an object at night. It is vastly used in many domains like security, healthcare, process control, and surveillance especially in defense vehicles where visualization at night would also mandatory for checking. In this paper, a thermal imaging camera is proposed to only be felicitated in regular as well as automated vehicles for better identification of objects especially at night when visibility is very less. Due to the huge variation in grayscales and pseudo-coloring values in the thermal image, a fuzzy-based CNN [FCNN] model is proposed to be applied to identify the boundaries of the objects. In this technique, the correlation between the thermal images of the moving object and its types is proposed to be trained with the novel FCNN model. The framed methodology is not only implementable in benched marked video datasets but also applicable in real-life conditions on the ground experimental scenario on a live feed video streaming. The results significantly indicate the enhancement of visual capacity in the TIPT compared to normal visual technique.
Year
DOI
Venue
2022
10.1007/s11760-021-02118-7
Signal, Image and Video Processing
Keywords
DocType
Volume
Thermal imaging, Fuzzy model, CNN, Object detection, Jaccard index, Inter-connectivity
Journal
16
Issue
ISSN
Citations 
6
1863-1703
0
PageRank 
References 
Authors
0.34
6
2
Name
Order
Citations
PageRank
Sayantan Nath100.34
C. Mala210.71