Written in front:
First of all, thank you brothers for subscribing, which gives me the motivation to create. During the creative process, I will do my best to ensure the quality of the work. If you have any questions, you can send me a private message, and let us work together to create brilliance.
Although the road is long, you will get there if you go; although the task is difficult, you will succeed if you do it. As long as you have the ambition of Foolish Old Man to move mountains, the perseverance of dripping water to penetrate rocks, be down-to-earth, work hard, and take small steps to reach thousands of miles, you will surely be able to turn your grand goals into a beautiful reality.
The dataset looks like this:
1. Introduction
At present, oceans and lakes are becoming more and more polluted, and reports of underwater garbage threatening biodiversity are common. Underwater garbage, such as discarded fishing gear, packaging that has not been properly recycled, littered plastic bags and soda bottles, partially floats on the surface, but most sinks to the bottom. Studies show that hundreds of thousands of aquatic mammals die every year from pollution caused by underwater trash. Therefore, in order to protect the growth of underwater animals and plants and maintain the ecological balance of oceans and lakes, underwater garbage removal work is required. When removing garbage, you need to be able to distinguish garbage from animals and plants in a complex underwater environment. After mastering the distribution of garbage, you need to accurately locate and pick up underwater garbage. Usually these tasks are completed by scuba divers, but the underwater environment is complex and dangerous, which requires a lot of manpower and material resources. Therefore, using underwater robots to detect and remove garbage has become an effective way to remove underwater garbage.
2. Data set
This data set can be used to publish papers. After subscribing, you can obtain the detailed information of the data set and a free paper.
The details of the dataset are as follows:
Test set pictures: 1144
verification set pictures