KR-20260067901-A - Waste sorting system with improved waste location accuracy

Abstract

The present invention relates to a recycling waste sorting system with improved accuracy in identifying the location of waste, and more specifically, to a recycling waste sorting system comprising: a conveyor (600) for transporting recycling waste (10); at least one RGBD camera (100) for capturing an RGB image (21) and a depth information map (22) of a pile of recycling waste (10) at least one layer transported on the conveyor (600); an image processing module (200) for processing image information of the RGBD camera (100); and a classification module (300) for recognizing and classifying recycling waste (20) to be sorted by object using image information processed by the image processing module (200); wherein the recycling waste sorting system further comprises a location calculation module (400), and the location calculation module (400) comprises: a segmentation execution unit (410) for performing segmentation of each recognized recycling waste (11); and a waste location calculation unit (420) for calculating the center coordinates and distance of each recycling waste (11) that has performed the segmentation.

Inventors

• 이영진
• 김태욱

Assignees

• 주식회사 로보원

Dates

Publication Date

20260513

Application Date

20241106

Claims (3)

1. As a recycling waste sorting system, A conveyor (600) for transporting recycled waste (10); At least one RGBD camera (100) for capturing an RGB image (21) and a depth information map (22) of a pile of recyclable waste (10) at least one layer being transported on the conveyor (600); An image processing module (200) that processes image information of the above RGBD camera (100); and It includes a classification module (300) that recognizes and classifies recyclable waste (20) to be sorted by object using image information processed by the image processing module (200) above, and The above recycling waste sorting system additionally includes a location calculation module (400), and The above position calculation module (400) is, A segmentation performing unit (410) that performs segmentation of each recognized recyclable waste (11); and A recycling waste sorting system with improved accuracy in identifying the location of waste, characterized by including a waste location calculation unit (420) that calculates the center coordinates and distance of each recycling waste (11) that has undergone the above-mentioned segmentation.
2. In paragraph 1, The above position calculation module (400) is, A recycling waste sorting system with improved accuracy in identifying the location of waste, characterized by additionally including a robot picking position calculation unit (430) that calculates picking position information of a sorting robot (500) using information from the waste location calculation unit (420) above.
3. In either Article 1 or Article 2, The above classification module (300) is, A recycling waste sorting system with improved accuracy in locating waste, characterized by using artificial intelligence to recognize and classify recycling waste (11).

Description

Recycling waste sorting system with improved waste location accuracy The present invention relates to a waste sorting system with improved accuracy in locating waste, and more specifically, to a waste sorting system with improved accuracy in locating waste that can classify and sort recyclable waste more accurately by utilizing RGB images and depth information together of irregular and large-scale transported piles of recyclable waste. In modern times, a wide variety of items, including disposable products, are produced and used to enhance convenience in human life and work, resulting in the generation of a large amount of waste. Among waste materials, there are recyclable materials such as metal, plastic, glass bottles, and paper. Recently, there has been a trend to separate and recycle recyclable waste in order to reduce resource consumption, improve energy efficiency through recycling, and minimize environmental damage by decreasing the volume of waste. However, despite efforts to improve the recycling rate, it remains low because waste is not yet sorted accurately. Furthermore, most recycling centers currently sort collected waste manually, resulting in a significant proportion of labor costs and limited working hours, which leads to significantly reduced efficiency. To address these issues and improve the waste recycling rate, development is underway for recycling sorting or classification systems. As an example of this, there is a recycling waste sorting device using an artificial intelligence model disclosed in Public Patent Publication No. 10-2023-0084017 (hereinafter referred to as the 'prior art'). A recycling waste sorting device (100) using an artificial intelligence model of the prior art can, as shown in FIG. 1, capture an image of waste on a conveyor through at least one camera in a pre-set sorting area, analyze the captured image based on an artificial intelligence model to determine the type of each waste on the conveyor, analyze the captured image using an artificial intelligence model to determine the type of waste for each waste on the conveyor, and control a sorting module so that the waste on the conveyor is sorted into a first collection bin according to the type of waste. As such, conventional technology classifies waste using computer vision and robotic technologies, but in actual waste sorting facilities, the shape and material of the waste are widely diverse, and since waste piles move along conveyors in large quantities, there is a problem in that the classification accuracy using computer vision technology is very low. Furthermore, as waste is transported in such piles, not only is the size of each waste item not accurately segmented, but the distance from the robot equipped in the sorting system is also inconsistent depending on the waste. This makes it difficult for the robot to determine the precise picking position for each waste item, which may also cause difficulties in accurately separating or stably lifting the waste. Figure 1 is a recycling waste sorting device using an artificial intelligence model shown in the prior art. Figure 2 is an example of recycled waste (10) being transported from a waste sorting system. Figure 3 is a block diagram of a recycling waste sorting system utilizing an RGBD camera according to the present invention. Figure 4 is an experimental example of image information of an RGBD camera according to the present invention. FIG. 5 is a block diagram of an image processing module (200) according to the present invention. FIG. 6 is an example diagram of the distance calculation of waste using a depth information map (22) according to the present invention. FIG. 7 is an example diagram of setting a region of interest mask (40) according to the present invention. FIG. 8 is an example diagram of an edited RGB image extraction according to the present invention. FIG. 9 is a block diagram of a position calculation module (400) according to the present invention. FIG. 10 is an example diagram of segmentation execution according to the present invention. FIG. 11 is a flowchart of a recycling waste sorting method according to the present invention. FIG. 12 is a flowchart of an image processing step (S20) according to the present invention. FIG. 13 is a flowchart of the sorted waste location extraction step (S40) according to the present invention. The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described in detail below together with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but can be implemented in various different forms, and the embodiments provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the present invention is defined only by the scope of the claims. Furthermore, the terms used in this