Development Of A Cotton Mining Robot In Nanjing Agricultural University

Recently, the team of Wang Ling, an associate professor of engineering at Nanjing Agricultural University, has developed a robot that not only can pick cotton, but also quickly and accurately judge the grade of seed cotton. It will hopefully solve the problem of cotton picking quality from the source.

For farmers, picking cotton is a tough job.

Wang Ling introduced that the labour force invested by artificial cotton accounts for about 50% of the whole production process.

For example, in the Xinjiang Construction Corps, 7 million mu of cotton is planted, and the annual picking cost is nearly 400 million yuan.

Although mechanical picking can improve the efficiency, it is easy to cause good and bad.

In order to solve the problem of cotton picking efficiency and quality, Wang Ling and his team began to study cotton picking robots since 2005.

According to the introduction, the researchers first used the CCD digital camera to collect the front image of the three cotton seed cotton 12 to 1, 7 grades in the early, middle and late stages. The total image of the cotton seed was 350. After that, the seed cotton image was segmented from the complex background by technical treatment. The 14 texture characteristics reflecting the cotton whiteness, yellow and impurity content and 16 shape characteristics reflecting the size and structure of the cotton seed were used to classify and classify the seed cotton, and the recognition rates reached 85.3% and 88.4% respectively.

When picking cotton picking robot, binocular vision camera and image acquisition card are used to collect cotton images, and image processing is done to locate cotton coordinates.

After approaching the plant, the robot picks up through the rotation of the manipulator arm.