Fry counting internet-of-things equipment and fry condition statistical method

Abstract

The invention discloses a fry counting Internet of Things device and a fry condition statistical method, and belongs to the technical field of Internet of Things. The Internet of Things equipment comprises a box body and a power supply box, and also comprises a remotely connected statistical unit; the power supply box supplies power to the box body, an acquisition unit, an identification unit and a control unit are arranged on the box body, and the acquisition unit is arranged at the bottom in the box body and used for containing fry to be detected; the recognition unit is arranged at the top in the box body and is used for shooting the fry in the collection unit; the control unit is connected with the identification unit and is used for controlling the identification unit to work, acquiring fish fry images shot by the identification unit and outputting processing results of the statistical unit; and the statistical unit is a remote server, is connected with the control unit through the internet, and is used for processing the fry images and obtaining the number of live fries, the number of dead fries and the fry cycle. The Internet-of-things equipment is designed to achieve real-time collection of fry images, Internet transmission and deep learning and recognition of fries.

Description

technical field [0001] The invention belongs to the technical field of the Internet of Things, and in particular relates to an Internet of Things device for counting fish fry and a statistical method for fish fry. Background technique [0002] In China, large-scale modular farming is an inevitable choice to meet market demand. Traditional fry counting is based on manual statistics or empirical weighing. In recent years, artificial intelligence and deep learning have developed vigorously, and intelligence has been applied in more and more fields. The new fry counter will inevitably replace the traditional fry counting method and become the first choice for farmers in fry statistics. [0003] Fry individuals are very small, about 2-4.5cm in length. The unit price of fry is calculated by piece, and the price is related to the quality of the fry. The statistics of traditional fish fry need to be counted manually or estimated by weight. This greatly increases the time cost an...

AI Technical Summary

Problems solved by technology

However, the patent uses an industrial computer to process the collected static pictures, and a set of equipment is equipped with an industrial computer, which does not realize the shared centralized processing of Internet cloud technology; the patent uses a traditional counting model, does not use deep learning, and does not realize Internet remote GPU centralized deal with

Images