Skip to content (access key 's')
Logo of Technion
Logo of CS Department
Logo of CS4People

The Taub Faculty of Computer Science Events and Talks

Technologies for fish growers: from ornamental fish in greenhouses to edible fish in the open sea
event speaker icon
Boaz Zion
event date icon
Tuesday, 25.12.2007, 11:30
event location icon
Room 337-8 Taub Bld.
Development of technologies for agriculture is very complex. The products (fruits, vegetables, flowers, livestock etc) are non-uniform and susceptible, the sceneries (illumination, topography, geometry etc) are complex and vary with time, and the low cost of the products is a limiting factor to the cost effectiveness of the technology.

When the technologies deal with live fish, the constraints, requirements and limitations are even more acute due to two main reasons: underwater conditions may be quite limiting and fish are extremely susceptible to stress.

Four projects aiming to assist fish growers will be presented: 1. Ornamental fish fry counting by image processing. Ornamental fish farms in Israel practice daily counting of the spawned fry. Tens of thousands of fry are manually counted daily in the larger farms, for feed management and stock assessment. A system for counting a day-old ornamental fry has been developed and tested. It is based on image processing and its accuracy reaches 99%.
2. Determination of guppy fish gender and quality. Guppy growers and breeders base quality assessment of their fish on three types of criteria: physiological, shape and color. Shape related features are mainly wholeness of fins, their size and shape, body/tail ratio and body proportions. Color features are complex and vary from strain to strain. A computer vision system for gender-based and quality sorting of guppy fish is described.
3. Real-time underwater sorting of edible fish species. Common carp (Cyprinus carpio), St. Peter’s fish (Oreochromis sp.) and grey mullet (Mugil cephalus), were sorted according to species while swimming in pond water containing algae and suspended sediments. Overall species recognition accuracy was 98%.
4. Development of a “Virtual-Cage Culture” technology for sea ranching of fish. An advanced technology for ranching and harvesting stocked fish, using behavioral conditioning is suggested. No pixels, no algorithms, yet.