Recent status

The quality inspection of glutinous rice is an important business node for grain processing and grain storage enterprises. The accuracy of the inspection results is directly related to the economic interests and reputation of the enterprise. At present, the moldy and imperfect grains in the quality inspection of glutinous rice are all manually inspected, recorded and counted based on artificial senses by personnel. The existence speed is slow, the accuracy rate is low, the missed detection rate and the false detection rate are high, and long-term work fatigue. At the same time, there is a lot of uncertainty, and there are risks such as collusion among personnel. The efficiency of quality inspection is also difficult to meet the requirements of automated operations. It needs to be solved in a smarter and more efficient way.

Product Description

The equipment uses the latest AI vision detection technology and is equipped with 3 high-resolution cameras to analyze the attributes of the front and back sides of glutinous rice. Through the registration algorithm, the front and back sides of the glutinous rice are registered one by one, and their respective attributes are combined to obtain a synthesis. The properties of a whole piece of glutinous rice; use deep neural network to segment the sticky glutinous rice at instance level to easily deal with the sticking situation of glutinous rice; at the same time, open the cloud platform to remotely train samples of different customers to meet customer customized classification standards.

This machine replaces manual work, can work 7*24 hours, detect the quality of glutinous rice with high precision, detect broken rice, chalky grains, imperfect grains, and moisture in the glutinous rice in time, and find whether there are mildew, worms, impurities and other problems. It can be used for daily sampling inspection before and after glutinous rice production.

The glutinous rice quality detector can be connected to upstream and downstream production equipment according to the specific production needs of customers on site. The parts in contact with the equipment and samples are made of medical-grade materials, which are safe and hygienic, intelligent in design, simple to operate, and convenient to maintain.

Core technology

1.Automatic binarization: use deep neural network to segment the foreground and background of the image, smoothly segment the grain edge, and accurately locate the grain to be analyzed.

2.Adhesion material segmentation algorithm: deep neural network segments the adhering grains to form independent and complete grains, which are analyzed and classified.

3.Multi-attribute recognition: It adopts a lightweight neural network and integrates a semi-supervised multi-attribute learning method. The user can label a small number of samples of the grain to be analyzed, and then the data model can be updated to perform fast and high-precision analysis of the grain.

Adavantages of equipment

• 1. AI algorithm: high stability, adapting to the environment and background disturbance; different defect samples can be automatically identified after training
• Dataization: Independent database, save multiple samples, analyze non-good products, and retain history
• Multi-orientation: 360 ° comprehensive inside and outside the samples
• High precision: detection accuracy can be high
• Modularization, can flexibly increase or decrease the detection function according to customer actual needs
• Easy to operate: It is easy to operate and easy to maintain
• Safety: Medical grade material manufacturing, fully compliant with medical supplies production environment