Object Detection on Thermal Images: Performance of YOLOv4 Trained on Small Datasets

Chaverot, Maxence; Carré, Maxime; Jourlin, Michel; Bensrhair, Abdelaziz; Grisel, R.

doi:10.14428/esann/2021.es2021-130

Cited by 7 publications

(7 citation statements)

References 10 publications

(11 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These categories are used to assess the performance of the detection model across different object sizes. The baseline score is obtained with the fine-tuned model presented in [28] on our validation subset. The four presented methods are: MTH (Multiscale Top-Hat), MTH LIP, UM (Unsharp Masking deblurring), SR (Super Resolution).…”

Section: Resultsmentioning

confidence: 99%

“…Thus, we compute multiple Top-Hat transforms with structuring elements of increasing size, the maximal radius being empirically chosen from detection performance and time constraints. Chaverot et al [28] proposed a method consisting of executing the neural network object detector on all these Top-Hat transform iterations. The next step is to concatenate the obtained results.…”

Section: Mathematical Morphology and Top-hat Lipmentioning

confidence: 99%

See 1 more Smart Citation

Improvement of small objects detection in thermal images

Chaverot

Carré²,

Jourlin

et al. 2023

ICA

View full text Add to dashboard Cite

Thermal images are widely used for various applications such as safety, surveillance, and Advanced Driver Assistance Systems (ADAS). However, these images typically have low contrast, blurred aspect, and low resolution, making it difficult to detect distant and small-sized objects. To address these issues, this paper explores various preprocessing algorithms to improve the performance of already trained object detection networks. Specifically, mathematical morphology is used to favor the detection of small bright objects, while deblurring and super-resolution techniques are employed to enhance the image quality. The Logarithmic Image Processing (LIP) framework is chosen to perform mathematical morphology, as it is consistent with the Human Visual System. The efficacy of the proposed algorithms is evaluated on the FLIR dataset, with a sub-base focused on images containing distant objects. The mean Average-Precision (mAP) score is computed to objectively evaluate the results, showing a significant improvement in the detection of small objects in thermal images using CNNs such as YOLOv4 and EfficientDet.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Mathematical Morphology and Top-hat Lipmentioning

confidence: 99%

Improvement of small objects detection in thermal images

Chaverot

Carré²,

Jourlin

et al. 2023

ICA

View full text Add to dashboard Cite

show abstract

“…We utilized the YOLO v4 deep convolutional neural network [ 26 ] pretrained on the MSCOCO dataset as our base object detector. This network seems to perform well on the LWIR images even without fine-tuning [ 27 ].…”

Section: Object Detection For Parametrized Spectrum Imagesmentioning

confidence: 99%

Bias-Tunable Quantum Well Infrared Photodetector

Biswal,

Yakimov,

Tokranov

et al. 2024

Nanomaterials

View full text Add to dashboard Cite

With the rapid advancement of Artificial Intelligence-driven object recognition, the development of cognitive tunable imaging sensors has become a critically important field. In this paper, we demonstrate an infrared (IR) sensor with spectral tunability controlled by the applied bias between the long-wave and mid-wave IR spectral regions. The sensor is a Quantum Well Infrared Photodetector (QWIP) containing asymmetrically doped double QWs where the external electric field alters the electron population in the wells and hence spectral responsivity. The design rules are obtained by calculating the electronic transition energies for symmetric and antisymmetric double-QW states using a Schrödinger–Poisson solver. The sensor is grown and characterized aiming detection in mid-wave (~5 µm) to long-wave IR (~8 µm) spectral ranges. The structure is grown using molecular beam epitaxy (MBE) and contains 25 periods of coupled double GaAs QWs and Al0.38Ga0.62As barriers. One of the QWs in the pair is modulation-doped to provide asymmetry in potential. The QWIPs are tested with blackbody radiation and FTIR down to 77 K. As a result, the ratio of the responsivities of the two bands at about 5.5 and 8 µm is controlled over an order of magnitude demonstrating tunability between MWIR and LWIR spectral regions. Separate experiments using parameterized image transformations of wideband LWIR imagery are performed to lay the framework for utilizing tunable QWIP sensors in object recognition applications.

show abstract

“…2 Early works of automatic target detection systems focused on approaches that could be broken into five subsections: pre-processing, classification/recognition, detection, feature selection, and feature extraction, [2][3][4][5] all methods of which require constant supervision from humans to extract the desired features by hand, which greatly slowed processing of the system. Recent research on automatic target detection has transitioned to focus on machine learning based approaches using modern techniques such as advanced sensors, [6][7][8][9] deep learning, [10][11][12] convolutional neural networks, 2, 13, 14 and more recently vision transformers. 15 There are two sensors commonly used to capture information for ATD/R systems: Synthetic Aperture Radars (SAR) and Infrared (IR).…”

Section: Introductionmentioning

confidence: 99%

“…Deep learning networks utilized dense and complex architectures that required tremendous amounts of computational power to function. 6,10,11 These models achieve high accuracy metrics, but are less then preferred when designing tiny and efficient systems.…”

Section: Introductionmentioning

confidence: 99%

Automatic target detection utilizing an edge IR vision transformer (EIR-ViT)

Adams

Depoian

Kurz

et al. 2023

Automatic Target Recognition XXXIII

View full text Add to dashboard Cite

The detection and recognition of targets within imagery and video analysis is vital for military and commercial applications. The development of infrared sensor devices for tactical aviation systems imagery has increased the performance of target detection. Due to the advancements of infrared sensors capabilities, their use for field operations such as visual operations (visops) or reconnaissance missions that take place in a variety of operational environments have become paramount. Many techniques implemented stretch back to 1970, but were limited due to computational power. The AI industry has recently been able to bridge the gap between traditional signal processing tools and machine learning. Current state of the art target detection and recognition algorithms are too bloated to be applied for on ground or aerial mission reconnaissance. Therefore, this paper proposes Edge IR Vision Transformer (EIR-ViT), a novel algorithm for automatic target detection utilizing infrared images that is lightweight and operates on the edge for easier deployability.

show abstract

Object Detection on Thermal Images: Performance of YOLOv4 Trained on Small Datasets

Cited by 7 publications

References 10 publications

Improvement of small objects detection in thermal images

Improvement of small objects detection in thermal images

Bias-Tunable Quantum Well Infrared Photodetector

Automatic target detection utilizing an edge IR vision transformer (EIR-ViT)

Contact Info

Product

Resources

About