CN212442146U

CN212442146U - Household garbage recognition system

Info

Publication number: CN212442146U
Application number: CN202021787931.8U
Authority: CN
Inventors: 李建涛; 杨建红; 黄文景; 房怀英; 张宝裕; 林伟端; 郑皎; 秦瑞聪
Original assignee: Huaqiao University
Current assignee: Huaqiao University
Priority date: 2020-08-24
Filing date: 2020-08-24
Publication date: 2021-02-02
Anticipated expiration: 2030-08-24

Abstract

The utility model provides a domestic waste identification system, include industrial computer, rubbish conveyor and follow colour camera, 3D camera and high spectrum camera that rubbish conveyor's direction of delivery arranged in proper order, colour camera's side below be provided with colour camera complex linear light source, the 3D camera other be provided with 3D camera complex laser emitter, be provided with under the high spectrum camera with high spectrum camera complex dome light source, rubbish conveyor colour camera the 3D camera with high spectrum camera respectively with the industrial computer communication is connected. The utility model discloses a domestic waste multidimension degree characteristic information is gathered to the mode that 3D camera, color camera and high spectrum camera combined, can show the discernment accuracy that improves domestic waste through many characteristic information fusion and level-to-level discernment for machine intelligence letter sorting can be applicable to the domestic waste letter sorting, helps improving letter sorting efficiency and improves letter sorting operating mode.

Description

Household garbage recognition system

Technical Field

The utility model relates to a refuse handling system, especially a household garbage recognition system.

Background

With the rapid increase of social economy, the consumption level of people is continuously improved. However, with the consumption of people, hundreds of millions of domestic wastes are brought. The domestic garbage is difficult to recover and can only be buried or burnt, so that the environment is polluted and more land resources are occupied.

With the coming of national policies and the improvement of environmental awareness of people, how to better recycle domestic garbage and change waste into valuables becomes a focus of attention of governments and enterprises. The most important ring of retrieving domestic waste is the letter sorting, and present domestic waste letter sorting mode mainly uses manual sorting as the owner, sorts inefficiency, and letter sorting operating mode is abominable simultaneously. In the face of billions of tons of household garbage, machine intelligent sorting is necessary to become a future trend instead of manual sorting, and the problem to be solved by machine intelligent sorting is how to identify the household garbage.

In view of the above, the applicant has made intensive studies to solve the above problems and has made the present invention.

SUMMERY OF THE UTILITY MODEL

An object of the utility model is to provide a help improving letter sorting efficiency and improving the domestic waste identification system of letter sorting operating mode.

In order to achieve the above purpose, the utility model adopts the following technical scheme:

the utility model provides a household garbage recognition system, includes industrial computer, rubbish conveyor and is located rubbish conveyor top is followed colour camera, 3D camera and the high spectrum camera that rubbish conveyor's direction of delivery arranged in proper order, rubbish conveyor colour camera the 3D camera with the high spectrum camera respectively with the industrial computer communication is connected.

As an improvement of the utility model, a linear light source matched with the color camera is arranged below the side of the color camera, a laser emitter matched with the 3D camera is arranged beside the 3D camera, a dome light source matched with the hyperspectral camera is arranged under the hyperspectral camera,

as an improvement of the utility model, install the halogen lamp on the dome light source.

As an improvement of the utility model, be provided with on the rubbish conveyor and be used for tiling rubbish dispersion devices on the rubbish conveyor, dispersion devices is located color camera keeps away from one side of 3D camera.

As an improvement of the utility model, the industrial computer has an image algorithm processing module.

Adopt above-mentioned technical scheme, the utility model discloses following beneficial effect has:

the utility model discloses a domestic waste multidimension degree characteristic information is gathered to the mode that 3D camera, color camera and high spectrum camera combined, and then can show the discernment accuracy that improves domestic waste through multi-feature information fusion and level-to-level discernment for machine intelligence letter sorting can be applicable to the domestic waste letter sorting, helps improving letter sorting efficiency and improves letter sorting operating mode.

Drawings

Fig. 1 is a schematic structural view of the household garbage recognition system of the present invention;

fig. 2 is the image algorithm processing flow schematic diagram of the household garbage recognition system of the utility model.

The designations in the figures correspond to the following:

10-a waste conveying device; 20-a color camera;

21-a linear light source; a 30-3D camera;

31-a laser emitter; 40-a hyperspectral camera;

41-dome light source; 42-halogen lamp.

Detailed Description

The invention will be further described with reference to the accompanying drawings and specific embodiments.

As shown in fig. 1, the present embodiment provides a household garbage recognition system, which includes an industrial personal computer (not shown in the figure), a garbage conveying device 10, and a color camera 20, a 3D camera 30 and a hyperspectral camera 40 that are located above the garbage conveying device 10 and are sequentially arranged along a conveying direction of the garbage conveying device 10, wherein the industrial personal computer has an image algorithm processing module, and is mainly used for performing multi-feature fusion on image information acquired by multiple cameras and obtaining location information and category of a material by combining with an algorithm, and such industrial personal computers with the image algorithm processing module can be directly purchased from the market, and certainly, if necessary, software needs to be installed on the industrial personal computer to implement a required control processing function.

The garbage conveying device 10 is used for conveying household garbage as a material, and when the garbage conveying device 10 is used, a hopper for dumping garbage is arranged at the head end of the garbage conveying device 10, and a sorting device or a manipulator for sorting garbage is arranged at the tail end of the garbage conveying device (the hopper and the sorting device or the manipulator are not part of the embodiment). The waste conveyor 10 may be a conventional belt conveyor, which is a commercially available product directly with the color camera 20, the 3D camera 30 and the hyperspectral camera 40, as well as the linear light source 21, the laser emitter 31, the dome light source 41 and the halogen lamp 42, which will be mentioned below.

The shooting direction of the color camera 20 is vertically downward, and the color camera is used for collecting color information and space information of the materials. The linear light source 21 matched with the color camera 20 is arranged below the color camera 20, the irradiation range of the linear light source 21 comprises the shooting range of the color camera on the conveying surface of the garbage conveying device 10, and the stability of image characteristic information acquisition of the color camera 20 is ensured by uniform light emitted by the linear light source 21.

The 3D camera 30 is used for collecting height information of the material, and a laser emitter 31 matched with the 3D camera 30 is arranged beside the 3D camera 30, and specifically, the 3D camera and the laser emitter 31 are respectively installed on the same horizontally arranged connecting plate. The light beam of the laser emitter 31 is emitted in a vertical direction and is located within the shooting range of the 3D camera 30 on the conveying surface of the waste conveyor 10. When in use, the 3D camera 30 obtains height profile information through laser emitted by the material.

The hyperspectral camera 40 is used for collecting spectral information and spatial information of materials, a dome light source 41 matched with the hyperspectral camera 40 is arranged right below the hyperspectral camera, the dome light source 41 is preferably a dome light source, and a halogen lamp 42 is arranged on the dome light source 41. The center of the dome light source 41 is provided with a through hole, and the hyperspectral camera 40 is located right above the through hole and the shooting direction of the hyperspectral camera is vertically downward. Therefore, the stability of the collection of the material reflection spectrum can be ensured through diffuse reflection.

The specific fixed connection structure of the color camera 20, the 3D camera 30, the hyperspectral camera 40, the linear light source 21, the laser emitter 31, the dome light source 41 and the halogen lamp 42 may be a conventional structure, for example, a fixed connection is realized by a support frame fixedly connected to the frame of the garbage transporting device 10, or a fixed connection is realized on a support frame fixed relative to the position of the garbage transporting device 10, and details thereof are not described herein. In addition, the garbage conveying device 10, the color camera 20, the 3D camera 30 and the hyperspectral camera 40 are respectively in communication connection with an industrial personal computer so as to receive and send signals or information, and a specific communication connection structure can be a conventional structure, such as a wire connection or a bluetooth wireless connection.

Preferably, a dispersing device (not shown) for spreading the garbage on the garbage conveying device 10 is arranged on the garbage conveying device 10 to avoid the garbage stacking from affecting the shooting. The dispersing device is located on the side of the color camera 20 away from the 3D camera 30, which may be a dispersing device commonly used in the garbage disposal field, and is not the focus of the embodiment and will not be described in detail here.

When the method is used, the image algorithm processing module screens dark materials of household garbage through image information provided by the color camera 20 and the 3D camera 30, and obtains position and category information of non-dark materials through image information provided by the hyperspectral camera 40 and the 3D camera 30, a specific processing method can be an existing method, and preferably, the embodiment provides a method with relatively high processing efficiency and accuracy in consideration of the fact that the existing method is relatively low in processing efficiency and accuracy, and specifically, referring to fig. 1 and 2, after the image algorithm processing module receives the image information sent by the color camera 20, the 3D camera 30 and the hyperspectral camera 40, the image algorithm processing module adopts the following steps to realize processing of the algorithm by taking the image information collected by the polyphaser as multiple features of the materials:

s1, matching the height image collected by the 3D camera 30 with the color image collected by the color phase 20 after threshold segmentation and scaling to obtain a color image, and sequentially carrying out image graying, image filtering, image binarization, edge detection and contour analysis on the color image to obtain the mass center and the rotation angle of the material so as to obtain the position information of the material. Because the detected object has obvious height characteristics, and the interferents in the detection environment are dust impurities with almost no thickness and no obvious height characteristics, the interferents can be well filtered through the height image, and further, the interferents can be well filtered through the color image matched with the height image.

S2, converting the RGB model (i.e. the three primary color models of red, green and blue) of the color map into HSV model (hue, saturation and brightness model), equalizing the HSV model, removing noise, and then distinguishing dark objects from non-dark objects by detecting color, wherein dark objects are usually not sorted but directly discarded due to relatively low added value, and thus are not further identified.

S3, importing the hyperspectral image information of the non-dark object acquired by the hyperspectral camera 40 into a prediction model for classification prediction to obtain the category information of the non-dark object, wherein the construction method of the prediction model comprises the following steps:

and S3.1, extracting 28 effective features from the wavelength features of the hyperspectral image information through wavelet transformation. The model needs to be trained to obtain the parameter weight firstly for image information prediction, because the spectral dimension of the hyperspectral data has 224 wavelength characteristics, direct training of the hyperspectral data is not only long in data but also time-consuming in training, and the embodiment extracts 28 effective characteristics from the hyperspectral data through wavelet transformation to train and predict, so that the problems are effectively avoided.

And S3.2, importing the hyperspectral image information (namely the hyperspectral image data after the characteristics are extracted) after the step S3.1 is finished into a model of an extreme learning machine for training to obtain the trained parameter weight.

And S3.3, putting the parameter weight into a pre-constructed model to obtain a prediction model.

After the construction of the prediction model is completed, the method can be applied to online real-time detection, and at the moment, the collected hyperspectral image information of the object which is not in the dark color is imported into the prediction model, so that the model can obtain the category information of the object.

S4, selecting three spectral features in the hyperspectral image information to obtain a three-channel pseudo-color image similar to the color image, and then matching the pseudo-color image with the height image after threshold segmentation and scaling (similar to the step S1) to obtain the position information and the height features of the non-dark object.

And sorting the materials by the sorting equipment or the mechanical arm according to the position information and the height characteristics. The embodiment mainly collects the rich image information of the objects through the multiple cameras, wherein the rich image information comprises color information, space information, height information and hyperspectral information, and the dark objects with low additional values can be well detected and positioned through algorithm processing, and the dark objects are accurately classified and positioned, so that the accuracy of subsequent sorting is improved.

The present invention has been described in detail with reference to the accompanying drawings, but the embodiments of the present invention are not limited to the above embodiments, and those skilled in the art can make various modifications to the present invention according to the prior art, and these all belong to the protection scope of the present invention.

Claims

1. The household garbage recognition system is characterized by comprising an industrial personal computer, a garbage conveying device and a high spectrum camera, wherein the high spectrum camera, the 3D camera and the high spectrum camera are sequentially arranged in the conveying direction of the garbage conveying device, and are positioned above the garbage conveying device, and the garbage conveying device is characterized in that the color camera, the 3D camera and the high spectrum camera are respectively in communication connection with the industrial personal computer.

2. The household garbage recognition system as claimed in claim 1, wherein a linear light source matched with the color camera is arranged below the side of the color camera, a laser emitter matched with the 3D camera is arranged beside the 3D camera, and a dome light source matched with the hyperspectral camera is arranged right below the hyperspectral camera.

3. The household garbage recognition system of claim 2, wherein a halogen lamp is mounted on the dome light source.

4. The household garbage recognition system as claimed in claim 1, wherein a dispersing device for spreading the garbage on the garbage conveying device is disposed on the garbage conveying device, and the dispersing device is located on a side of the color camera away from the 3D camera.

5. The household garbage recognition system as claimed in any one of claims 1-4, wherein the industrial personal computer has an image algorithm processing module.