Quantifying plastic refuse in water area helps to understand how plastic refuse accumulates in water area and is essential for targeted cleanup efforts. Currently, the most common methods for quantifying plastic in water area are human visual counting and sampling using nets, but such methods are costly and labor-intensive. This study proposes a watershed refuse identification algorithm based on an improved YOLOv4. Lightweight improvements to YOLOv4. EfficientNetB1 is used to replace the backbone network of YOLOv4, and the Depthwise Convolution is used to replace the original convolution to reduce the number of model parameters and computation. The anchors are re-clustered using k-means algorithm to improve the accuracy. The experimental results show that the improved algorithm improves the detection speed by 11.2% and reduces the number of parameters by 76.54% compared with YOLOv4 at the expense of 0.69% recognition accuracy.

With the progress of science and technology and the continuous improvement of living standards, robots are more and more widely used in life and production, and robot grasping technology is also constantly improving. In practical application, accurate grasp detection of target objects is an important part of robot grasping task. In this paper, the parallel two-fingered gripper is used as the end of the robot arm's grasping, and the research status of grasping detection, which is the key part in the grasping process of the robot arm based on vision, is summarized. The 2D planar grasping and 6-DOF spatial grasping are compared and analyzed in detail. At the same time, it also summarizes the commonly used evaluation indexes of capturing data sets and capturing detection, and points out the challenges faced by vision-based robot capturing and the future direction of solving these challenges.

With the continuous development of society and the deepening of the level of science and technology, the fields explored by human beings are becoming more and more extensive, and the requirements for exploration are becoming more and more detailed. However, due to the limitation of mobility, most of the existing amphibious robots still can't be used to complete the specified tasks in complex environments. In order to meet the reliability of motion in complex environments, referring to ePaddle mechanism can enhance the mobility of amphibious robots. Based on the new movement mechanism of eccentric paddle mechanism (ePaddle), according to its principle and prototype, this paper proposes a compound wheeled amphibious robot, which realizes the movement of the robot through an improved eccentric lever mechanism. This mechanism includes multiple gaits, and can adapt to various complex terrains by actively changing the position of the paddle shaft to follow the alternate movement of the wheel foot and the paddle shaft. The key feature of the improved mechanism is to design the rotation of the planet carrier to drive the blades to expand and contract to realize underwater movement. This paper mainly introduces the kinematics analysis and prototype design, and simulates various gait movements through simulation software to verify the feasibility of the new design.