Random Forest - a curated list of resources regarding random forest
In this work, we present an extensive description and evaluation of our method for blood vessel segmentation in fundus images based on a discriminatively trained, fully connected conditional random field model. Standard segmentation priors such as a Potts model or total variation usually fail when dealing with thin and elongated structures. We overcome this difficulty by using a conditional random field model with more expressive potentials, taking advantage of recent results enabling inference of fully connected models almost in real-time. Parameters of the method are learned automatically using a structured output support vector machine, a supervised technique widely used for structured prediction in a number of machine learning applications. Our method, trained with state of the art features, is evaluated both quantitatively and qualitatively on four publicly available data sets: DRIVE, STARE, CHASEDB1 and HRF. Additionally, a quantitative comparison with respect to other strategies is included. The experimental results show that this approach outperforms other techniques when evaluated in terms of sensitivity, F1-score, G-mean and Matthews correlation coefficient. Additionally, it was observed that the fully connected model is able to better distinguish the desired structures than the local neighborhood based approach. Results suggest that this method is suitable for the task of segmenting elongated structures, a feature that can be exploited to contribute with other medical and biological applications.
One of the first steps in automatic fundus image analysis is the segmentation of the retinal vasculature, which provides valuable information related to several diseases. In this work, we present an extensive description and evaluation of our method for blood vessel segmentation in fundus images based on a discriminatively trained, fully connected conditional random field model. This task remains a challenge largely due to the desired structures being thin and elongated, a setting that performs particularly poorly using standard segmentation priors, such as a Potts model or total variation. We overcome this difficulty by using a conditional random field model with more expressive potentials, taking advantage of recent results enabling inference of fully connected models almost in real-time. Parameters of the method are learned automatically using a structured output support vector machine, a supervised technique widely used for structured prediction in a number of machine learning applications. The evaluation of our method is performed both quantitatively and qualitatively on DRIVE, STARE, CHASEDB1 and HRF, showing its ability to deal with different types of images and outperforming other techniques, trained using state of the art features.
Almost in every image processing or analysis work, image pre-preprocessing is crucial step. In medical image analysis, pre-processing is a very important step because the further success or performance of the algorithm mostly dependent on pre-processed image. In this lab, we are working with 3D Brain MRI data. In case of working with brain MRI removing the noise and bias field (which is due to inhomogeneity of the magnetic field) is very important part of preprocessing of brain MRI. To do so, we widely used algorithm Anisotropic diffusion, isotropic diffusion which can diffuse in any direction, and Multiplicative intrinsic component optimization (MICO) have been used for noise removal and bias field correction respectfully. Both quantitative and qualitative performance of the algorithms also have been analyzed.
Intensity Inhomogeneity Correction - a tool for correcting intensity inhomogeneities (the "bias field" artifact) in MRI
Retrospective and Prospective Techniques of Illuminance Correction
line segment detector(lsd) & edge drawing line detector(edl) & hough line detector(standard and probabilistic)
Keeping roads in a good condition is vital to safe driving. To monitor the degradation of road conditions is one of the important component in transportation maintenance which is labor intensive and requires domain expertise. Automatic detection of road damage is an important task in transportation maintenance for driving safety assurance. The intensity of damage and complexity of the background, makes this process a challenging task. A deep-learning based methodology for damage detection is proposed in this project after being inspired by recent success on applying Deep- learning in Computer Sciences. A dataset of 9,053 images is taken with the help of a low cost smart phone and a quantitative evaluation is conducted, which in turn demonstrates that the superior damage detection performance using deep-learning methods perform extremely well when compared with features extracted with existing hand-craft methods. Using convolutional neural networks to train the damage detection model with our dataset, we use the state-of-the-art object detection method, and compute the accuracy and runtime speed on a GPU server. At the end, we show that the type of damage can be distinguished into eight types with acceptable accuracy by applying the proposed object detection method.
Matlab implementation of Mean-shifted Gray Pixel <Revisiting Gray Pixel for Statistical Illumination Estimation>VISSAP 2019
Diffusion anisotrope dans les images et detection de contours, 2017
