Review on Deep Learning Algorithms for Object Detection

Sasirekha, R.; Jeyshri, J.; TinaVictoria, A.; Subha, J.; Kamaleswari, P.

doi:10.1007/978-981-99-1726-6_32

Cited by 2 publications

(1 citation statement)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Performance is affected by neural network architecture and training data quality, which require lots of data and processing. Mystery image format noise model [16]. Beyond variational, quantum total variation Television lowers visual noise.…”

Section: Introductionmentioning

confidence: 99%

Hybrid total variance void-based noise removal in infrared images

Pandurangan,

Nagappan

2024

IJEECS

View full text Add to dashboard Cite

An artifact known as an image is what makes the depiction of a thing or a person feasible. An image is a representation of visual perception and has a physical appearance that is analogous to that of the subject being portrayed. In situations when there is insufficient illumination, such as at night or when there is a lot of background noise, the use of infrared imagery can help improve the accuracy of object detection. Infrared images are able to account for a wide variety of noises, including those that are the result of sensor faults, lens distortion, software artifacts, blur, and other problems. It is difficult to do qualitative and quantitative analysis on thermal images due to the significant levels of noise that are present in these images. Eliminating noise in an infrared image by employing the total variance void (TVV) denoising technique while preserving the integrity of the image’s boundaries and texture. Denoising thermal images make use of a technique that is both efficient and reliable thanks to an integrated algorithm that combines TV denoising and Noise2Void (N2V). Strengths of the two methods, it is possible to produce denoised images of superior quality with improved retention of edge and texture detail.

show abstract

Section: Introductionmentioning

confidence: 99%