This paper studies the development status of domestic and foreign cotton stalk harvester. On this basis, a double roller cotton stalk Harvester is designed, which is suitable for small plots. It can pull out the entire cotton stalk through two pairs of reverse rotation of the roller and the forward thrust of tractor. In order to increase the friction force between the roller and the cotton stalk, the two pairs of roller axes are designed with a certain angle and the surface is designed into a rugged, spiral into distribution, which can adapt to the different diameters of cotton stalk harvesting.

Topographic mapping and modeling is one of the most basic and important work in hydraulic engineering construction. This paper present a work flow which contains data measurement, pre-process and post-process with three dimensional laser scan technology. By adopting B-tree and its expansion structure, a multi-level point cloud management and storage method has been established. Then propose a contour generation method based on scan data and the equivalent point interpolation data. A self-adaptive mesh Delaunay algorithm is introduced with point cloud data in order to generate triangular irregular network digital elevation model. Finally, the modeling accuracy and error has been analysis, the results proves this method is with high efficient and accuracy and can be provide assistance for the project construction.

The study and application of machine vision technique for the detection of agricultural products, which includes the quality detection and grading of fruits, dried fruits, vegetables, are discussed. It summarizes the existing problems of detection and grading of agricultural products based on machine vision, and analyzes the methods of solving problems through establishing a national collaborative innovation center.

The development and application of the machine vision technology is greatly liberating the human labor force and improved the production automation level and the situation of human life, which has very broad application prospects. The intelligent empty bottle inspection robot this paper studies is a typical application of the machine vision in the industrial detection. This paper mainly introduces the concept of machine vision, some important technology related to automatic cleaning robots and application of the machine vision in the production of all areas of life.

In order to study the feasibility of the automatic diagnostician based on the images of palm, the general framework of automatic diagnostician is designed. It is implemented after analysis of the process of image recognition of palm images by computer and the main research content of each functional module is pointed out. Using the method of image processing, some palm images are collected by scanner. After proper preprocessing, images of palm's clear borderline can be acquired. The design of the philosophy system and images preprocessing have positive significance to the development of automatic diagnostician.