Intelligent sorting trash cans have gradually entered people's lives, and the scientific principles contained in them also have certain research value. The development and production process of its teaching projects not only allows students to experience the application of artificial intelligence technology, but also allows students to participate in the practical activities of garbage classification, effectively improving their innovative practical ability and problem-solving ability.

        This article further innovates and optimizes the "intelligent sorting trash can" project, and designs a fully automatic, low-cost intelligent sorting trash can.

In terms of material selection, we selected the artificial intelligence camera module (hereinafter referred to as the artificial intelligence camera) developed by the Kendryte K210 processor as the core as the image recognition device, and the Arduino UNO control board as the control core of the entire system, together with ultrasonic sensors and steering gear Modules, infrared sensors, LED light strips, MP3 modules, etc. work together to complete the work. The sensors of each part work in coordination and cooperation to realize the function of fully automatic intelligent garbage classification, and also have a certain "human-bucket" interaction function (as shown in Figure 1).

Innovative design of the overall function

This project innovates the design of intelligent sorting trash cans from five aspects: automatic sensing, image recognition, fully automatic delivery, overflow judgment and voice broadcast. The workflow of the project is shown in Figure 2.

  Figure 2. Working flow chart of automatic intelligent sorting garbage can

1. Automatic induction

In order to solve the problem of the peculiar smell of the trash can, we added an infrared sensor door to the opening of the trash can, and use the infrared sensor to judge whether someone is approaching. When the user is close to the trash can, the infrared sensor senses the person, the steering gear turns, drives the sensor door, opens the trash can lid, and the user can put in the garbage. Among them, the tray above the trash can is used for users to place trash. Compared with the traditional foot-opening trash can, the automatic sensor opening can bring users a better experience.

2. Image recognition

We replace the picture garbage with physical garbage or "three-dimensional garbage" made by 3D printing to simulate real usage scenarios to the greatest extent. When we use artificial intelligence cameras for image recognition, we increase the process of model training, that is, first collect images of all garbage, and then mark the data set. After multiple iterations of training, the artificial camera can automatically identify the type of garbage. In teaching, if teachers can guide students to explore and experience the whole process of computer image recognition in this way, students will definitely realize the importance of data to results and further master the principles of image recognition.

3. Fully automatic delivery

In terms of the putting function of the smart trash can, we innovatively use two steering gears to control the left and right and up and down movements of the tray: by moving left and right, the garbage is carried above the corresponding garbage collection box; by moving up and down to dump the garbage, it automatically, Dispose of garbage accurately. Fully automatic delivery is the most important function of this work, and it is also the biggest innovation point. It has preliminarily realized the fully automatic process from "throwing" to "classification" to "precise delivery" of individual garbage. If similar functions are added to the teaching project, it will not only innovate the function of the project, but also better meet the actual needs, and help students further appreciate the convenience brought by artificial intelligence technology to life.

4. Overfill judgment

Ultrasonic sensors built into the bottom of the tray provide "overfill" status. The sensor is able to determine whether each bin is overflowing before each time the user puts in the waste, that is, when the tray is moved over the corresponding bin. If a garbage collection box is full and the user puts the same type of garbage into the garbage can again, the system will stop "putting" and give a voice prompt.

5. Voice broadcast

After the intelligent sorting trash can recognizes the garbage, it will voice broadcast the recognition result, including the name and type of the garbage, so that users can confirm whether the garbage they put in is correctly identified. If a garbage collection box is overflowing, the intelligent sorting garbage can will light up with a red LED light, accompanied by a voice broadcast, reminding the user to empty the garbage can in time to ensure the normal operation of the system.