Computer Vision for Microscopy Applications

Citations Over TimeTop 10% of 2007 papers

Abstract

A simulation is a reproduction under controlled circumstances of a real-life situation. The term has recently become strongly associated with numeric evaluation of a computer model due to the increase in speed and availability of computing resources. This increase in speed has led to much interest in stochastic simulation, where processes with elements that are random are simulated. An attractive branch of stochastic simulation termed Monte Carlo simulation uses deterministic models driven by stochastic input sequences to approximate the distributions of output variables over time. To do this, a good deterministic model of the process is needed in addition to a good method of generating realistic input sequences. Correct input sequences are a prerequisite for reliable results from stochastic simulation. To generate them, the modeller must either generate input sequences by hand, develop a model based on intuition or understanding of the process, or use existing data. Generating input sequences by hand is a tedious and error-prone process and intuition is not a particularly verifiable source of information. This means that data-driven model development has been gaining favour steadily as data becomes more accessible. This chapter covers three aspects of input signal generation: First, the basic theory of Markov processes and hidden Markov models is reviewed with a view on using them as generating processes for input models. Second, signal segmentation is introduced. This is the first step in identifying state transition probabilities for discrete Markov processes. In this part, novel work done on the identification of state transitions using multi-objective optimisation is introduced and ideas for future research are posed. Third, the problem of estimating state transition probabilities from the segmented signals is discussed, touching on the issues that modellers should be aware of. Markov processes have featured strongly in stochastic sequence identification and generation for many years, but some of the related problems are still active research fields.

Related Papers

• → An Object Detection and Pose Estimation Approach for Position Based Visual Servoing(2017)5 cited
• → Self-monitoring to improve robustness of 3D object tracking for robotics(2011)4 cited
• → 3-D tracking of a moving object by an active stereo vision system(2002)9 cited
• → Robust object tracking based on RGB-D camera(2014)
• → DETERMINING QUALITY REQUIREMENTS AT THE UNIVERSITIES TO IMPROVE THE QUALITY OF EDUCATION(2018)