番茄采摘机器人快速无损作业研究(英文版)

Chapter 1 History and Present Situations of Robotic Harvesting Technology: A Review 1.1 An Industry of Fresh-Eat Fruits and Vegetables and Its Labor-Cost Harvesting Fruits and vegetables are both daily necessities， and also they are important economiccrops. According to the statistics， the global production of total fruits and vegetablesin 2019 reached 8.83 x l08 t and 11.30 x l08 t， respectively.Globally， the rate betweenfresh-eat and processed fruits and vegetables is about 7:3.Chinese vegetable and fruitplanting area and output both rank the first in the world， but the proportion ofprocessing fruits and vegetables is only about 5%. Usually， it is not necessary for the harvesting of processing fruits and vegetablesto distinguish the ripeness， and also a certain damage is tolerant in the harvest. Forexample， tomato fruit can be whole-plant harvested and apple fruit can be harvestedmechanically by vibratory excitation.ln developed countries， the non-selective mech-anized harvesting of processed fruits and vegetables has been gradually popularized.But for the larger proportion of fresh-eat fruits and vegetables， the non-selectivemechanized harvesting method cannot adapt to both the individual difference of thefruit maturity and the harsh demand of non-destructive harvest. So far， it is stilldependent on human labor for selective harvesting. With the gradual mechaniza-tion of the production of fruit and vegetable cultivation， harvesting has become thelast link to break through the whole process of mechanical operation. According tothe investigation， the labor consumption of strawberry production in Japan reaches20，000 h/ha [11， and the harvest takes up about 40% of the totallabor amount [1，2].Meanwhile， the shortage of agricultural labor and the rising cost oflabor haveseriously affected the development of the fruit and vegetable industry. In China， inrecent years， the labor force， especially the young and middle-aged labor force， hasalso been rapidly transferred to other industries. In the busy farming season， the laborshortage has begun to appear in the vast rural areas. The labor intensity of the elderlyand women in rural areas has greatly increased， and the production efficiency hasdecreased obviously. The contradiction between the rapid development of fruit and vegetable produc-tion， the shortage of agricultural labor， and the excessive intensity of labor isbecoming more and more obvious， and the replacement of complex manual selectiveharvesting can only be realized through the in-depth study of the technology of theharvesting robot. The research and development of fruit and vegetable harvestingrobot are of great significance for reducing the labor intensity of agricultural practi-tioners，liberating the agriculturallabor force andimproving the intensive productionlevel of fruits and vegetables. 1.2 The History and Current Situation of the Development of Robotic Harvesting Equipment in the Whole World A typical harvesting robot for fruits or vegetables is usually composed of mobileplatform， manipulator， end-effector， vision system， and control system. Since fruitand vegetable species and varieties， and cultivation patterns are all numerous andcomplicated， various kinds of harvesting robots and their end-effectors have beendeveloped at home and abroad. The action principle， structure form， complexity，operation effect， and performance also have a very big difference. 1.2.1 Tomato Harvesting Robots 1. Fresh-eat tomato and its robotic harvesting problem As a favorite fresh-eat vegetable， its robotic harvesting has been paid much attentionby researchers worldwide. Concerned research has been camed out continuously formany years， and a series of achievements have been produced. At the same time， the tomato is also one of the fruits and vegetables that are mostdifficult to be harvested by robots. At present， in the face of fresh food， commontomato fruits are usually picked as single fruit one by one， while cherry-tomatofruit is usually picked in clusters. Compared with cucumber， eggplant， apple， andother fruits and vegetables， there is usually 3-5 tomato fruit in one cluster. Theygrow densely and touch each other， and the difference of fruit-stem posture is moresignificant (Fig. 1.1). The great difference of growth posture and distribution posesa greater challenge to the implementation ofintelligent robotic harvesting: (1) Recognition of the target fruit The close and occlusion of the fruit are more serious. For the vision system of theharvesting robot， although the color difference between mature tomato fruit andleaves is distinct， it is difficult to identify and locate the target fruit since multiplefruit images are integrated into one or even completely overlapped to be difficult tobe segmented [3， 4]. Fig. 1.1 Difference of growth posture and distribution among the fruit of tomato， cucumber， and Eggplant (a) Tomato (b) Cucumber (c) Eggplan