The Guxin ultra-large‑particle ore intelligent sorter is specially designed for 3–20 cm lump raw ore, operating in both dry and wet modes. It is compatible with nearly a hundred types of metallic and non‑metallic ores, including quartz, gypsum, limestone, calcium carbonate, fluorite, barite, coal gangue, copper oxide, hematite, phosphate rock, and more. Equipped with a 5,400‑pixel high‑speed CCD and AI deep‑learning algorithms, it accurately detects micron‑level color differences and surface textures, achieving a sorting accuracy of up to 99.9%. Featuring a tracked conveyor system, it ensures stable operation with minimal crushing. With a processing capacity of 30–100 tons per hour and a waste‑carry‑over ratio of ≤2%, it significantly boosts efficiency while reducing losses. Key components include a 5,400‑pixel full‑color CCD, multispectral imaging, AI deep‑learning algorithms, a reinforced double‑layer track, and an automated high‑speed pneumatic valve.

Keywords：Gu Xin Ultra-Large-Particle Ore Color Sorter

The Guxin ultra-large‑particle ore color sorter is a specialized sorting device designed for primary screening at the mine’s front end. It focuses on oversized raw ore ranging from 30 to 200 mm (3–20 cm), addressing the longstanding challenges of low efficiency, high costs, and inconsistent quality associated with traditional manual sorting. Featuring pre‑sorting of large waste rock, enrichment of high‑grade ore, dual‑mode operation for both dry and wet conditions, and stable performance under extreme operating environments, it serves as a core component of intelligent mining systems—reducing burdens at the front end while boosting productivity downstream. Key specifications include a 5,400‑pixel full‑color CCD, multispectral imaging, AI‑driven deep‑learning algorithms, a reinforced double‑layer conveyor belt, and an air‑jet‑based high‑speed valve system.

I. Application Scenarios

● Open-pit/Underground Raw Ore Primary Sorting: Open‑pit, shaft‑head, and stockpile sorting of large‑size run‑of‑mine ore (3–20 cm) enables the direct removal of waste rock, interbedded gangue, and discolored material, reducing crushing, grinding, and flotation loads by more than 30%.  
● Bulk purification of non-metallic minerals: Quartzite/Quartz Ore: Large‑size raw ore is processed to remove yellow skin, black spots, and gangue, thereby increasing the grade of the quartz sand feedstock. Calcite/Calcium Carbonate: Bulk material is graded and impurities removed to supply heavy calcium carbonate powder and sheet‑producing applications. Kaolin/Coal Gangue: Bulk gangue is separated from kaolin, boosting the recovery rate of kaolin. Fluorite/Barite/Dolomite: Large‑size ore is screened to eliminate waste rock, enhancing the grade of the concentrate.  
● Large-scale enrichment of metallic ores: Tungsten/Molybdenum ores: Large‑size run‑of‑mine ore is screened to remove waste rock, thereby increasing the grade fed to the mill. Gold/Silver ores: Large waste rock is discharged in advance to minimize gold loss. Hematite/Magnetite: High‑grade lump ore is sorted according to color and texture.  
● Building Materials and Recycled Resources: Granite/Marble rough blocks: Remove discolored or heavily cracked large pieces. Pebbles/Landscape stones: Sort large pieces, remove impurities, and select shapes. Solid waste aggregates: Separate usable material from large chunks of construction rubble.

II. Technical Parameters

Product model	KS4-1350	KS4-1650	KS4-1950
Number of channels	512	640	768
Production (t)	30-60	45-80	60-100
Purity rate (%)	≥99	≥99	≥99
Power supply voltage (V/Hz)	220/380 (50 Hz)	220/380 (50 Hz)	220/380 (50 Hz)
Power (kW)	7	8	9
Air supply pressure (MPa)	0.5-0.8	0.5-0.8	0.5-0.8
Air consumption (L/min)	<4.5	<4.5	<4.5
Machine weight (kg)	2450	2600	2830
External dimensions (Length × Width × Height) (mm)	4907*1820*2727	4907*2120*2727	4907*2420*2727

Note: The above production figures are based on color‑sorted pebbles measuring 3–15 cm, with 5% impurities.

III. Classification by Ore Type

Non-metallic minerals: Quartz, potassium feldspar, calcite, calcium carbonate, kaolin, coal gangue, fluorite, barite, dolomite, granite, marble, pebbles, perlite, limestone, talc, fused quartz, glass, wollastonite, volcanic rock, salt deposits, bauxite, brucite, garnet, vermiculite, phosphate rock, gypsum, wollastonite, magnesite, aggregates, expansive soils, Hanyang white marble, chicken-blood stone, rain‑flower stones, brick and concrete, metallic silicon, and others.  
Metallic ores: Tungsten ore, molybdenum ore, gold ore, silver ore, hematite, copper oxide, manganese ore, lead–zinc ore, and others.

IV. Particle Size Range for Sorting

Standard large-grain model: 30mm–200mm (3–20 cm), covering extremely large blocks of raw ore. Fine-grained adaptation: 30–80 mm (3–8 cm): Most commonly used, suitable for mainstream large‑size crushed ore. 80–150 mm (8–15 cm): Medium‑sized raw ore. 150–200 mm (15–20 cm): Extra‑large raw or rough ore. Compatible with both dry and wet processing: suitable for both dry and wet separation; no drying required for wet separation.

V. Sorting and Recognition Dimensions

● Color sorting: Micron-level color‑difference detection eliminates yellow‑skinned, black‑spotted, and heavily discolored ore chunks. Texture sorting identifies differences between gangue and mineral textures, separating interlayered rock and waste material. Shape sorting detects cracks, chipping, and deformed pieces, removing non‑conforming material.
● Internal Impurity Identification: Multispectral imaging penetrates the surface layer to detect internal interlayers and foreign inclusions. Automated grading of mixed ores: large‑size mixed material containing different ore types and grades is automatically sorted and graded.

VI. Core Competitive Advantages

● Exclusive design for extra-large particles: The only intelligent sorter that stably handles oversized ore pieces measuring 5–20 cm, filling a critical gap in the intelligent processing of large‑size raw ore; its frame, material distributor, and pneumatic valves have all been reinforced to withstand the impact of heavy‑weight materials.  
●AI multispectral high-definition recognition: Accuracy exceeds 99.9%, featuring a high-speed, full-color 5,400-pixel CCD and multispectral imaging to capture subtle color differences and textures in large ore samples. Powered by AI deep learning, it builds models tailored to local bulk ore specimens, becoming increasingly precise with use and remaining unaffected by lighting, dust, or humidity. Sorting accuracy is ≥99.9%, with false‑reject and false‑accept rates below 0.1%; carry‑over is ≤2%, significantly lower than the industry average of 5%.  
● Compatible with a wide range of extreme operating conditions: High stability: One‑button switching between dry and wet modes enables direct sorting of moist or water‑washed large‑size ore, eliminating the need for drying. Enhanced dust‑proof and waterproof design ensures reliable performance in high‑dust, high‑clay, and extreme temperature environments (–25°C to 40°C). It supports 24‑hour continuous operation with long mean time between failures, making it ideal for high‑intensity mining operations.  
● High processing capacity + low carryover ratio: High efficiency and low loss, with a throughput of 30–80 tons per hour (depending on particle size and ore type), delivering more than 20 times the productivity of manual sorting. A dual‑layer conveyor belt ensures uniform material distribution, combined with free‑fall separation, preventing large pieces from overlapping or getting stuck and achieving precise, one‑pass sorting. Waste‑to‑ore carry‑over is ≤2%, minimizing valuable mineral loss and boosting recovery rates by 5–10 percentage points.  
● Labor substitution + cost reduction and efficiency enhancement: One piece of equipment replaces 30–50 manual sorters, addressing challenges such as labor shortages, high labor costs (saving millions annually), and inconsistent sorting standards. Its modular design ensures easy disassembly, assembly, and maintenance; with universally compatible spare parts, operational and maintenance costs are reduced by more than 50%. The system features low overall power consumption (≤7.5 kW), delivering energy efficiency that is over 30% better than comparable equipment.

VII. Value Delivered to the Customer

● Front-end workload reduction, back-end efficiency enhancement: By pre‑removing more than 30% of waste rock, equipment wear and electricity costs for crushing, grinding, and flotation are reduced, resulting in an overall reduction of over 20% in mineral processing costs.  
● Enhance concentrate grade and selling price: Large pieces of waste rock and impure ore are discharged in advance, boosting concentrate grade by 5–15 percentage points and delivering a product premium of over 30%.  
● Improve mineral recovery rate: Reduce the loss of large, high-grade ore blocks, increase metal recovery rates by 5–10 percentage points, and conserve resources to prevent waste.  
● Intelligent automation replacing manual labor: Fully automated operation reduces reliance on manual labor and lowers management costs; standardized sorting criteria ensure consistent product quality.  
● Fast to commission, compact footprint: Installation and commissioning are straightforward, with production commencing within 7 days; the equipment occupies only about 10 m², making it well-suited to compact mine sites.

VIII. Case Studies of Color Sorting for Certain Ores

9. Color-sorting sites of some customers