IPDB › China Industry › Manufacturing & Processing Machinery › Sorting Machine
Double Layer Cobblestone Color Sorter for Separate Cobblestone Sorting High Purity Silicon1
Sorting Machine, Cobblestone Color Sorter, Ore Color Sorter
Hefei Mingde Optoelectronic Technology Co.,Ltd
Preview:
Product DescriptionProduct DescriptionMingde Optoelectronics' AI sorting machine innovatively integrates deep convolutional neural networks (CNN) with visible light sorting technology. Through its independently developed multi-dimensional feature extraction algorithm, it achieves precise identification and efficient sorting of complex ores. This device adopts the unique local connection, weight sharing and multi-convolutional kernel architecture of CNN. During the training process, it independently builds a multi-dimensional feature database of materials, and the sorting accuracy significantly exceeds that of traditional photoelectric sorting machine.Working principle1. High-precision multispectral/hyperspectral sensing: The equipment is equipped with advanced optical sensors that can accurately capture subtle spectral feature differences on the surface and inside of the ore (such as color, texture, mineral composition, impurity distribution, etc.), with a resolution far exceeding traditional manual or mechanical sorting methods.2. AI-driven intelligent recognition algorithm: The core algorithm based on deep learning can process massive spectral data in real time and accurately establish a mapping model between ore characteristics and target minerals/impurities. The system has strong self-learning and adaptive capabilities and can cope with natural fluctuations in ore properties.3. High-speed real-time processing and precise blowing: The powerful data processing platform ensures that when the ore passes at high speed (the processing capacity can reach tens to hundreds of tons/hour), the identification and decision-making are completed instantly, and the high-pressure gas valve array is driven to spray and separate the target particles with millisecond-level precision.4. Modular and customized design: The equipment adopts a modular design, and the core optical module, sorting execution module, control module, etc. can be flexibly configured according to different ore characteristics (particle size, density, sorting requirements) to meet the diverse needs from coarse-grained pre-throwing to fine-grained selection.Technical ParametersProductMineral particle size(mm)Capacity(t/h)Air pressure(mpa)Sorting accuracy(%)Optimal take-out ratioPower(kw)Dimension(mm)Weight(kg)MAI-D-25-103-40.559610:132945*1390*1640100010-305-60.630-6015-230.65MAI-D-45-106-80.559610:154680*2015*1700150010-3010-120.630-6030-450.65MAI-D-65-109-120.559610:175200*2502*1925350010-3015-180.630-6045-680.65MAI-TD-630-8045-800.69610:1106790*2502*2190600080-150150-2000.65MAI-S-45-107-90.5599.820:195155*3430*2680350010-3012-140.630-6035-500.65Technical Advantages:1,Empowered by AI Deep Convolutional Neural NetworksThe CNN artificial intelligence algorithm is applied to the field of optoelectronic sorting, and the intelligent analysis of material image features is achieved through a multi-layer neural network architecture, breaking through the bottleneck of traditional color sorting technology.2,Multi-dimensional Stereoscopic Feature RecognitionThe innovatively developed AI photoelectric sorting technology can automatically extract multi-dimensional features such as texture, shape, color, quality and luster of materials, greatly expanding its application scenarios and meeting the personalized sorting needs of different sorting scenarios and material types.3,Real-time Dynamic Acceleration TechnologyBy adopting a collaborative optimization scheme of model compression and hardware, the inference speed of CNN is increased to the millisecond level, effectively guaranteeing the requirements of industrial-grade real-time sorting.4,Small sample learning systemIntegrating the transfer learning framework with industrial-grade image enhancement technology, it can still reach a high recognition accuracy rate in scenarios with a sample less than a thousand cases.5,High-Performance Computing PlatformIt is equipped with a gigabit Ethernet industrial camera and a multi-GPU parallel computing platform. The computing platform adopts CNN for material type discrimination technology, which can accurately analyze the small subtle surface features of materials.Ores can be sortedCertificationsFAQ1.Can we visit your factory?2.How long about the machine guarantee period?Yes,our factory is located in Hefei city,Anhui Province, 2.5 hour bullet train distance from Shanghai,warmly welcome your visiting.One year. And we supply lifelong software upgrade services for our customers.3.Can you test samples to check sorting effect?Yes,sample sending for testing is welcomed.
Purchasing Agent
Note: Send your message to supplier or manufacturer.
Welcome to our Professional and comprehensive procurement services...
Message:
OK ! We suggest you detail your product requirements and company information here.
Enter between 20 to 4,000 characters.
Your inquiry content must be between 20 to 4000 characters.