To address the issue that existing natural radiation-based coal and gangue identification methods cannot directly calculate the gangue mixing rate at discharge outlets, a monitoring system for coal discharge quantity in fully mechanized top coal caving based on a binocular camera has been designed. This system aims to calculate the dilution ratio by monitoring the discharge volume at the coal discharge port and combining it with the quality data of gangue provided by the natural radiation identification method, thus assisting in determining the appropriate timing for closing the discharge port. The system employs a grating projection method to project a grid image of the coal flow surface transported by the scraper conveyor near the discharge port. It utilizes a binocular camera to scan the grating and obtain high-precision point cloud data of the coal flow profile. Through projection and micro-mesh partitioning, an organized three-dimensional point cloud model of the coal flow is constructed. A finite element volume calculation method is introduced for slicing the model, and the trapezoidal rule is used to calculate the volume of each slice, leading to the overall volume of the coal flow through summation. Additionally, a telescopic support structure has been designed to adapt to different field requirements, allowing for flexible adjustment and safety protection of the detector. Laboratory experiments for coal quantity monitoring demonstrate that the system maintains a measurement error of less than 5% under various coal flow accumulation conditions, indicating high accuracy and good stability. Specifically, under flat and piled conditions, the measurement errors are controlled within 5% and between 4.6% and 4.9%, respectively, with a root mean square error of 0.2952. The coal discharge quantity monitoring system designed in this study is effective and can provide reliable real-time monitoring data for the intelligent coal discharge in fully mechanized mining faces, thereby supporting the determination of the dilution ratio. This has significant implications for improving the accuracy and practicality of automatic coal gangue identification technology. The successful implementation of this system provides a new technical means for the intelligent construction of coal mines and has broad application prospects.