Image Background Noise Impact on Convolutional Neural Network Training

Rajnoha, Martin; Bürget, Radim; Povoda, Lukáš

doi:10.1109/icumt.2018.8631242

Cited by 8 publications

(6 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this study, marker‐based forceps tracking is performed to remove the background information, followed by an CNN‐based posture estimation. The tracking image helps to accelerate training convergence and enables posture estimation, without the influence of the background areas. Another advantage of the marker‐based method is that it does not require that the endoscope is able observe the entire instrument.…”

Section: The Proposed Methodsmentioning

confidence: 99%

Three‐dimensional posture estimation of robot forceps using endoscope with convolutional neural network

Mikada

Kanno

Kawase

et al. 2020

Robotics Computer Surgery

View full text Add to dashboard Cite

Background In recent years, there has been significant developments in surgical robots. Image‐based sensing of surgical instruments, without the use of electric sensors, are preferred for easily washable robots. Methods We propose a method to estimate the three‐dimensional posture of the tip of the forceps tip by using an endoscopic image. A convolutional neural network (CNN) receives the image of the tracked markers attached to the forceps as an input and outputs the posture of the forceps. Results The posture estimation results showed that the posture estimated from the image followed the electrical sensor. The estimated results of the external force calculated based on the posture also followed the measured values. Conclusion The method which estimates the forceps posture from the image using CNN is effective. The mean absolute error of the estimated external force is smaller than the human detection limit.

show abstract

Section: The Proposed Methodsmentioning

confidence: 99%

Three‐dimensional posture estimation of robot forceps using endoscope with convolutional neural network

Mikada

Kanno

Kawase

et al. 2020

Robotics Computer Surgery

View full text Add to dashboard Cite

show abstract

“…If the model is trained using the original data set described above, the model convergence process is affected by the fact that the region of interest covers most of the back-bottom noise, which is the case of low signal-to-noise ratio (SNR) (more than 50% of the image is background), as shown in Figure 2. M Rajnoha et al [23] proposed that redundant background (noise) in images can be subtracted to aid training if the available information covers a limited area. Experimental results demonstrate that removing background noise can greatly help the training process and speed up convergence in the case of low SNR.…”

Section: Roi Acquisition and Enhancementmentioning

confidence: 99%

“…The signal enhancement method we devised differs from replacing the subtracted background with black pixels [23]. Based on the prior information of the reflection area of the object in the ROI, a two-dimensional masking weight matrix is designed using the idea of normal distribution to suppress background noise.…”

Section: Roi Acquisition and Enhancementmentioning

confidence: 99%

See 1 more Smart Citation

Vehicle classification applying many‐to‐one input network architecture in 77‐GHz FMCW radar

Jiang

Zhang

Wang

et al. 2021

IET Radar Sonar & Navi

View full text Add to dashboard Cite

In this paper, we proposed a Many-to-One Input Network Architecture (MOINA) for the classification of similar structured vehicles (bus, truck and car). The inputs of the architecture are the multiple-masked region-of-interest of the same detected vehicle from Range-Doppler maps, which are acquired by FMCW radar. The proposed method is trained with a supervised system yielding a classification accuracy of 98%. MOINA shows good classification performance in a practical situation. Besides, the F1-score of buses, trucks and cars are 98.7%, 98.0% and 97.6%, respectively.

show abstract

“…The process of transfer learning is accomplished either by reusing features from the second to Image noise has been a primary concern in computer vision tasks. The presence of image noise, in form of a redundant background, crucially affects the outcome of image analysis [10]. CNNs, at the cost of resources, efficiently classify images that may be affected when the region of interest is significantly smaller.…”

Section: Introductionmentioning

confidence: 99%

Impacts of Background Removal on Convolutional Neural Networks for Plant Disease Classification In-Situ

Yin

et al. 2021

Agriculture

View full text Add to dashboard Cite

Convolutional neural networks have an immense impact on computer vision tasks. However, the accuracy of convolutional neural networks on a dataset is tremendously affected when images within the dataset highly vary. Test images of plant leaves are usually taken in situ. These images, apart from the region of interest, contain unwanted parts of plants, soil, rocks, and/or human body parts. Segmentation helps isolate the target region and a deep convolutional neural network classifies images precisely. Therefore, we combined edge and morphological based segmentation, background subtraction, and the convolutional neural network to help improve accuracy on image sets with images containing clean and cluttered backgrounds. In the proposed system, segmentation was applied to first extract leaf images in the foreground. Several images contained a leaf of interest interposed between unfavorable foregrounds and backgrounds. Background subtraction was implemented to remove the foreground image followed by segmentation to obtain the region of interest. Finally, the images were classified by a pre-trained classification network. The experimental results on two, four, and eight classes of datasets show that the proposed method achieves 98.7%, 96.7%, and 93.57% accuracy by fine-tuned DenseNet121, InceptionV3, and DenseNet121 models, respectively, on a clean dataset. For two class datasets, the accuracy obtained was about 12% higher for a dataset with images taken in the homogeneous background compared to that of a dataset with testing images with a cluttered background. Results also suggest that image sets with clean backgrounds tend to start training with higher accuracy and converge faster.

show abstract

Image Background Noise Impact on Convolutional Neural Network Training

Cited by 8 publications

References 9 publications

Three‐dimensional posture estimation of robot forceps using endoscope with convolutional neural network

Three‐dimensional posture estimation of robot forceps using endoscope with convolutional neural network

Vehicle classification applying many‐to‐one input network architecture in 77‐GHz FMCW radar

Impacts of Background Removal on Convolutional Neural Networks for Plant Disease Classification In-Situ

Contact Info

Product

Resources

About