Nanjing Agricultural University developed a cotton picking robot

Recently, the team of Wang Ling, an associate professor of the School of Engineering at Nanjing Agricultural University, developed a robot that can not only pick cotton, but also quickly and accurately determine the grade of seed cotton, which is expected to solve the problem of cotton picking quality at the source.

For farmers, collecting cotton is a chore. Wang Ling introduced that the labour input by artificial cotton picking accounts for about 50% of the entire production process. For example, in the Xinjiang Construction Corps, planting 7 million acres of cotton, spending nearly 400 million yuan each year to pick flowers. Although the efficiency of mechanical picking can be improved a lot, it can easily lead to the advantages and disadvantages. In order to solve the problem of cotton picking efficiency and quality, since 2005, Wang Ling and the team began to study cotton picking robots.

According to reports, the researchers first used CCD digital cameras to collect 350 images of the frontal, middle, and late “Summer 12” grade 1-7 seed cotton, and then processed the seed cotton image from a complex background through technical processing. Using the 14 texture features that reflect the whiteness, yellowness, and impurity content of the seed cotton and the 16 shape features that reflect the size and structure of the cotton pad, the maturity and classification of the seed cotton were identified. The recognition rates reached 85.3% and 88.4 respectively. %. When picking the actual picking robot, the binocular vision camera and image capture card were used to collect the cotton image, and image processing was performed to locate the cotton coordinate position. After approaching the plant, the robot picks by turning the robotic joints.