关键词: 名优茶, 智能采摘, 图像分割, 嫩芽识别, 采摘点定位

Abstract: At present, the tender bud picking of highquality tea is still in the stage of manual picking. Labor shortage and lack of machinery have become the limiting factors for expanding production of highquality tea. As a laborintensive industry, tea production is in urgent need of standardization and mechanization. Whether it is bulk tea or famous tea, mechanization picking is needed to meet the growing production demand. With the development of computer vision, artificial intelligence, automatic control and other new technologies, intelligent agricultural machines have been applied to all aspects of agricultural production. This paper comprehensively analyzes the latest progress in research on key technologies for intelligent picking of highquality tea in our country focuses on the introduction of tender buds image segmentation technology research, tender bud recognition and picking point location technology research, and points out that the current research has problems such as poor robustness of segmentation algorithms, low identification and positioning accuracy, and nonintegration of agricultural machinery and agronomy. In order to improve the precision of picking, the methods such as improved algorithms, technological innovation, integration of agricultural machinery and agronomy are proposed. The aim is to provide a theoretical reference for the realization of intelligent picking of highquality tea.

Key words: high-quality tea, intelligent picking, image segmentation, tender , bud recognition, picking point localization