Water quality and organism dispersion are critical factors that can influence results when operating the XperCount and therefore require special attention. Below are a set of guidelines to follow to avoid common issues. We also provide a few examples of images that meet the preferred criteria for best results.
1. Remove all foreign particles
Some particles are easier to detect than others. Carbon pellets, for example, are easily distinguished from living organisms and can be removed by hand. Other forms of debris, such as artemia particles or hair are difficult to see by eye, but are equally as important to remove using the proper filters. When processed by the camera, these foreign objects become more apparent; if too many exist in the sample, you risk lowering the accuracy of your counting session.
Food particles (dark dots)
2. Always use clean water
Dark or heavily tinted water may reduce the camera’s ability to detect organisms in the bucket. Tank water typically contains high amounts of feed and waste particulates that must be removed from the sample to ensure accurate results.
Dark water with algae or additives
Dirty tank water
3. Always use living organisms
The XperCount2 uses advanced recognition algorithms and can detect when multiple organisms are overlapping or crossed. When using live organisms, the animals swim freely in the bucket and create an even dispersion, reducing the amount of overlap. Dead organisms tend to form large clumps at the bottom of the bucket and cause severe clustering. The camera is unable to differentiate individual animals in severe clustering situations and will negatively impact the accuracy of your count and size data.
Dead organisms causing clustering
4. Always use healthy organisms
It’s always best to use organisms that have not been sitting outside of their normal living conditions for an extended period of time. Stressful conditions, such as lack of oxygen or feed, can cause cannibalism and impact color. Cannibalism pollutes the water and can cause clustering due to weak organisms.
Cannibalization causing clustering and half organisms
Good sampling examples
The two images below feature clean water and healthy organisms which naturally disperse inside the bucket, providing an ideal basis for optimal counting and sizing accuracy!
Good image #1
Good image #2