Abstract: As the core equipment of gravity beneficiation, the stability of the discharge state of the jigs directly affects the sorting efficiency, product quality, and resource recovery rate. Propose a method for monitoring the automatic discharge status of jigs with binocular vision support. Firstly, establish a binocular vision detection model and improve the accuracy of target detection through binocular vision camera calibration to obtain jigs images; Secondly, combining guided filtering and single scale Retinex model to perform filtering processing on the images of the jigs, eliminating the blurring phenomenon in the images; Finally, the filtered image is input into the global structural information learning network module to mine the prior information of the jig image. The LSTM gating mechanism is used to fully learn the discharge status information of the jig, thereby determining the automatic discharge status of the jig and achieving state monitoring. The experimental results show that the proposed method has good image filtering effect and high state monitoring accuracy.