@ Jaimee Ong : ...using an infrared camera and to observe the affect of spoilage in said produce.

Using an infrared camera, an infrared (IR) image is a snapshot of detected reflected heat energy in the infrared spectrum for a collection of physical shapes in a non-reflected sunlight scene. Examples of a pair of infrared images showing a nighttime spruce tree are given in the first of the attached images captured by @Braden Cross.

The second pair of attached images show the triangulation of the pair of infrared images. The main advantage in triangulating infrared images is that it provides a means of measure irregular shapes that have been recorded by anfrared camera. In the study of fish, this would be extremely useful, since minute changes in the fish could be recorded in conditions where there is an absence of sunlight.

The third pair of attached images show Alexandroff nerve complexes contained in the triangulated images. Such nerve complexes were introduced by Alexandroff in 1935. See, for example, page 39 in

P. Alexandroff

Elementary Concepts of Topology

https://www.maths.ed.ac.uk/~v1ranick/papers/aleks.pdf

Dear Jaimee,

Non-destructive testing is the best for your work.

The easiest way is to first find a niche or classes of fish, then speak to fish ichthyologists to get fundamental information about fish freshness, tissues, etc. or basics of their tissues, if you will, for the purpose you are trying to design the experiment.

If your method would be non-destructive, would be much simpler rather than to catch fishes and perform destructive tests and analytics on them.

The application of which, as in your attached paper says, is efficiency and automation, rather than to perform manual and time-consuming processes.

Good project!

Best regards,

J.

https://arxiv.org/pdf/1805.10106

Preprint Underwater Fish Species Classification using Convolutional N...

https://www.mikechatzidakis.com/home/2017/7/30/using-convolutional-neural-networks-to-identify-fish-species-in-camera-footage

http://courses.washington.edu/fish538/resources/Allken%20et%20al%202018.pdf

https://www.researchgate.net/post/finding_good_image_database_for_classification_of_fish_images_using_tensorflow

https://www.researchgate.net/publication/305674437_FishAPP_A_mobile_App_to_detect_fish_falsification_through_image_processing_and_machine_learning_techniques

https://www.researchgate.net/publication/303711963_Fish_species_classification_in_unconstrained_underwater_environments_based_on_deep_learning_Fish_classification_based_on_deep_learning

http://ceur-ws.org/Vol-1866/paper_166.pdf

https://ieeexplore.ieee.org/document/8124505/

Rather than (i) killing all fish at a time and (ii) killing one each day, there can be two ways.

1) for non-destructive methods the method suggested by J. Rafiee are good

OR

2) for studying the decay effects of fish - you can kill a batch of fishes each day

This is an important question. In addition to what has already been observed, there is a bit more to add.

First, I agree with @ J. Rafiee : Non-destructive testing is the best for your work.

Second, from an image processing (and computer vision) perspective, recording your observations with a cell phone or video camera or a webcam would provide you with a fairly complete history of what is happening to the fish in your experiments. Notice, also, that an infrared camera can be set up so that you can record what is happening to the fish either during daylight hours or at night.

Going a step further (after you acquire a video), triangulate or tessellate each video frame. This will make it possible to measure minute changes within each polygon covering the fish recorded in a video frame. If you convert your video frames to the HSV (hue saturation value) color space, then you can also measure the hue angle as well as well as the wavelength of fish pixels. This will help you quantify your observations.

For more about this, see

Book Foundations of Computer Vision

James F Peters & J. Rafiee

Good Day gentlemen. but the cope of my research is to correlate the wavelet transformation coefficients to the actual Quality assessment of food. Which in my case is Microbial analysis. Any suggestions on where the killing of my samples would have to kept in ice in the first day. or keep them alive and kill them off one by one until the desired days is reached.

Or else you have an idea already that the correlation of microbial analysis to image processing values do coincide with one another. When one is increasing the resulting other does increase as well.

I value the addition of information that can be modified into my research. Loving the idea onto using an infrared camera and to observe the affect of spoilage in said produce. Thank you.

@ Jaimee Ong : ...using an infrared camera and to observe the affect of spoilage in said produce.

Using an infrared camera, an infrared (IR) image is a snapshot of detected reflected heat energy in the infrared spectrum for a collection of physical shapes in a non-reflected sunlight scene. Examples of a pair of infrared images showing a nighttime spruce tree are given in the first of the attached images captured by @Braden Cross.

The second pair of attached images show the triangulation of the pair of infrared images. The main advantage in triangulating infrared images is that it provides a means of measure irregular shapes that have been recorded by anfrared camera. In the study of fish, this would be extremely useful, since minute changes in the fish could be recorded in conditions where there is an absence of sunlight.

The third pair of attached images show Alexandroff nerve complexes contained in the triangulated images. Such nerve complexes were introduced by Alexandroff in 1935. See, for example, page 39 in

P. Alexandroff

Elementary Concepts of Topology

https://www.maths.ed.ac.uk/~v1ranick/papers/aleks.pdf