Design and experimental study of the end-effector for broccoli harvesting

The end-effector is an important part of the broccoli harvesting robot.Aiming at the physical characteristics of a large broccoli head and thick stem,a spherical cutting tool broccoli harvesting end-effector was designed in this study.First,the physical characteristics of broccoli were tested,and physical parameters such as the broccoli head diameter and stem diameter of broccoli were measured.The maximum cutting force of broccoli stems under different cutting angles was tested.Second,according to the physical characteristics and harvesting process of broccoli,the end-effector was designed,and the mathematical model of kinematics and dynamics was established.Based on the results of dynamic analysis,the end-effector rod was optimized,and the unilateral width of the slider was 40 mm,the length of the connecting rod was 120 mm,and the length of the crank was 42 mm.The mechanism needed an external driving force of 140.54 N to cut the broccoli stem.Therefore,a 32 mm cylinder with a load rate of 50%was selected as the power source.Finally,the feasibility of the broccoli harvesting end-effector was verified by the harvesting test.Experiments showed that the overall harvesting success rate of the end-effector is 93.3%,and the smoothness rate of the stem section is 83.3%.The harvesting performance of the broccoli end-effector was verified.This lays a foundation for agricultural robots to harvest broccoli.

Identification of fruit and branch in natural scenes for citrus harvesting robot using machine vision and support vector machine

With the decrease of agricultural labor and the increase of production cost,the researches on citrus harvesting robot(CHR)have received more and more attention in recent years.For the success of robotic harvesting and the safety of robot,the identification of mature citrus fruit and obstacle is the priority of robotic harvesting.In this work,a machine vision system,which consisted of a color CCD camera and a computer,was developed to achieve these tasks.Images of citrus trees were captured under sunny and cloudy conditions.Due to varying degrees of lightness and position randomness of fruits and branches,red,green,and blue values of objects in these images are changed dramatically.The traditional threshold segmentation is not efficient to solve these problems.Multi-class support vector machine(SVM),which succeeds by morphological operation,was used to simultaneously segment the fruits and branches in this study.The recognition rate of citrus fruit was 92.4%,and the branch of which diameter was more than 5 pixels,could be recognized.The results showed that the algorithm could be used to detect the fruits and branches for CHR.

Design of structured-light vision system for tomato harvesting robot

In order to improve the operating precision of the harvesting robot,a vision system for intelligently identifying and locating the mature tomato was designed.The active detection method based on structured-light stereo vision was expected to deal with the problem of variable illumination and target occlusion in the glasshouse.The maximum between-cluster variances of hue(H)and saturation(S)value were adopted as the threshold for color segmentation,which weakened the impact on the image caused by the light intensity variation.Through the limit on the pixel size and circularity of the candidate areas,the vision system recognized the fruit area and removed the noise areas.The fruit’s 3D position was computed on the basis of spatial relationship between the laser plane and the camera,when the linear laser was projected on the centre area of the mature fruit.The blue view-scanning laser stripe pixels on the mature fruit were extracted according to its Cb color characteristic.As the field test results show,the measurement error on the fruit radius is less than 5 mm,the centre distance error between the fruit and camera is less than 7 mm,and the single axis coordinate error is less than 5.6 mm.This structured-light vision system could effectively identify and locate mature fruit.

Design and test of robotic harvesting system for cherry tomato

Harvesting of fresh-eating cherry tomato was highly costly on labor and time.In order to achieve mechanical harvesting for the fresh-eating tomato,a new harvesting robot was designed,which consisted of a stereo visual unit,an end-effector,manipulator,a fruit collector,and a railed vehicle.The robot configuration and workflow design focused on the special cultivating condition.Three key parts were introduced in detail:a railroad vehicle capably moving on both ground and rail was adopted as the robot’s carrier,a visual servo unit was used to identify and locate the mature fruits bunch,and the end-effector to hold and separate the fruit bunch was designed based on the stalk’s mechanical features.The field test of the new developed robot was conducted and the results were analyzed.The successful harvest rate of the robot was 83%,however,each successful harvest averagely needed 1.4 times attempt,and a single successful harvesting cycle cost 8 s excluding the time cost on moving.

A new strawberry harvesting robot for elevated-trough culture

In order to improve robotic harvesting and reduce production cost,a harvesting robot system for strawberry on the elevated-trough culture was designed.It was supposed to serve for sightseeing agriculture and technological education.Based on the sonar-camera sensor,an autonomous navigation system of the harvesting robot was built to move along the trough lines independently.The mature fruits were recognized according to the H(Hue)and S(Saturation)color feature and the picking-point were located by the binocular-vision unit.A nondestructive end-effector,used to suck the fruit,hold and cut the fruit-stem,was designed to prevent pericarp damage and disease infection.A joint-type industrial manipulator with six degrees-of-freedom(DOF)was utilized to carry the end-effector.The key points and time steps for the collision-free and rapid motion of manipulator were planned.Experimental results showed that all the 100 mature strawberry targets were recognized automatically in the harvesting test.The success harvesting rate was 86%,and the success harvesting operation cost 31.3 seconds on average,including a single harvest operation of 10 seconds.The average error for fruit location was less than 4.6 mm.

Development of a tomato harvesting robot used in greenhouse

A tomato harvesting robot was developed in this study,which consisted of a four-wheel independent steering system,a 5-DOF harvesting system,a navigation system,and a binocular stereo vision system.The four-wheel independent steering system was capable of providing a low-speed steering control of the robot based on Ackerman steering geometry.The proportional-integral-derivative(PID)algorithm was used in the laser navigation control system.The Otsu algorithm and the elliptic template method were used for the automatic recognition of ripe tomatoes,and obstacle avoidance strategies were proposed based on the C-space method.The maximum average absolute error between the set angle and the actual angle was about 0.14°,and the maximum standard deviation was about 0.04°.The laser navigation system was able to rapidly and accurately track the path,with the deviation being less than 8 cm.The load bearing capacity of the mechanical arm was about 1.5 kg.The success rate of the binocular vision system in the recognition of ripe tomatoes was 99.3%.When the distance was less than 600 mm,the positioning error was less than 10 mm.The time needed for recognition of ripe tomatoes and pitching was about 15 s per tomato,with a success rate of about 86%.This study provides some insights into the development and application of tomato harvesting robot used in the greenhouse.

Kinematics and trajectory planning of a cucumber harvesting robot manipulator

In order to reduce cucumber harvesting cost and improve economic benefits,a cucumber harvesting robot was developed.The cucumber harvesting robot consists of a vehicle,a 4-DOF articulated manipulator,an end-effector,an upper monitor,a vision system and four DC servo drive systems.The Kinematics of the cucumber harvesting robot manipulator was constructed using D-H coordinate frame model.And the inverse kinematics which provides a foundation for trajectory planning has been solved with inverse transform technique.The cycloidal motion,which has properties of continuity and zero velocity and acceleration at the ports of the bounded interval,was adopted as a feasible approach to plan trajectory in joint space of the cucumber harvesting robot manipulator.Moreover,hardware and software based on CAN-bus communication between the upper monitor and the joint controllers have been designed.Experimental results show that the upper monitor communicates with the four joint controllers efficiently by CAN-bus,and the integrated errors of four joint angles do not exceed four degrees.Probable factors resulting in the errors were analyzed and the corresponding solutions for improving precision are proposed.

Recognition and localization system of the robot for harvesting Hangzhou White Chrysanthemums

To realize the robotic harvesting of Hangzhou White Chrysanthemums,the quick recognition and 3D vision localization system for target Chrysanthemums was investigated in this study.The system was comprised of three main stages.Firstly,an end-effector and a simple freedom manipulator with three degrees were designed to meet the quality requirements of harvesting Hangzhou White Chrysanthemums.Secondly,a segmentation based on HSV color space was performed.A fast Fuzzy C-means(FCM)algorithm based on S component was proposed to extract the target image from irrelevant background.Thirdly,binocular stereo vision was used to acquire the target spatial information.According to the shape of Hangzhou White Chrysanthemums,the centroids of stamens were selected as feature points to match in the right and left images.The experimental results showed that the proposed method was able to recognize Hangzhou White Chrysanthemums with the accuracy of 85%.When the distance between target and baseline was 150-450 mm,the errors between the calculated and measured distance were less than 14 mm,which could meet the requirements of the localization accuracy of the harvesting robot.

Design and simulation of an integrated end-effector for picking kiwifruit by robot

The harvesting of fresh kiwifruit is a labor-intensive operation that accounts for more than 25%of annual production costs.Mechanized harvesting technologies are thus being developed to reduce labor requirements for harvesting kiwifruit.To improve the efficiency of a harvesting robot for picking kiwifruit,we designed an end-effector,which we report herein along with the results of tests to verify its operation.By using the established method of automated picking discussed in the literature and which is based on the characteristics of kiwifruit,we propose an automated method to pick kiwifruit that consists of separating the fruit from its stem on the tree.This method is experimentally verified by using it to pick clustered kiwifruit in a scaffolding canopy cultivation.In the experiment,the end-effector approaches a fruit from below and then envelops and grabs it with two bionic fingers.The fingers are then bent to separate the fruit from its stem.The grabbing,picking,and unloading processes are integrated,with automated picking and unloading performed using a connecting rod linkage following a trajectory model.The trajectory was analyzed and validated by using a simulation implemented in the software Automatic Dynamic Analysis of Mechanical Systems(ADAMS).In addition,a prototype of an end-effector was constructed,and its bionic fingers were equipped with fiber sensors to detect the best position for grabbing the kiwifruit and pressure sensors to ensure that the damage threshold was respected while picking.Tolerances for size and shape were incorporated by following a trajectory groove from grabbing and picking to unloading.The end-effector separates clustered kiwifruit and automatically grabs individual fruits.It takes on average 4–5 s to pick a single fruit,with a successful picking rate of 94.2%in an orchard test featuring 240 samples.This study shows the grabbing–picking–unloading robotic end-effector has significant potential to facilitate the harvesting of kiwifruit.

Nonlinear tracking control of a two link oil palm harvesting robot manipulator

Automated harvesting of oil palm trees requires research and development efforts in several robotics areas,including manipulator control.The objective of this paper was to apply nonlinear Lyapunov based control method for joint angles tracking of a two-link oil palm harvesting robot manipulator with uncertain system parameters.Four different controllers,including exact model knowledge,adaptive,sliding mode control and high gain feedback control were proposed and simulated.Stability analyses were performed for each case in the absence and presence of bounded disturbance.The controllers were then compared against each other based on their performances and control efforts.

Preprocessing method of night vision image application in apple harvesting robot

Due to the low working efficiency of apple harvesting robots,there is still a long way to go for commercialization.The machine performance and extended operating time are the two research aspects for improving efficiencies of harvesting robots,this study focused on the extended operating time and proposed a round-the-clock operation mode.Due to the influences of light,temperature,humidity,etc.,the working environment at night is relatively complex,and thus restricts the operating efficiency of the apple harvesting robot.Three different artificial light sources(incandescent lamp,fluorescent lamp,and LED lights)were selected for auxiliary light according to certain rules so that the apple night vision images could be captured.In addition,by color analysis,night and natural light images were compared to find out the color characteristics of the night vision images,and intuitive visual and difference image methods were used to analyze the noise characteristics.The results showed that the incandescent lamp is the best artificial auxiliary light for apple harvesting robots working at night,and the type of noise contained in apple night vision images is Gaussian noise mixed with some salt and pepper noise.The preprocessing method can provide a theoretical and technical reference for subsequent image processing.

Pneumatic bio-soft robot module:Structure,elongation and experiment

Soft robotics,often taking inspirations from biomimetics,is an exciting novel research field and has great capability to work with creatures.A new type of bio-soft robot inspired by elephant trunk and octopus was proposed,which has a promising application in robotic agricultural harvesting.The two modularized structures,basal segment and caudal segment,were elaborated in detail.They both have three side chambers and one central chamber,which are reinforced by springs for better stretch performance in axial direction without radial expansion.All the chambers can act as drivers when inflated with compressed air.More importantly,the central chamber was designed for regulating the stiffness of the robot module as needed in application.The primary static model for axial elongation was established for the fundamental analysis of the bio-soft robot module’s features,such as iso-force,isobaric and isometric characteristics.Simulation and experimental results showed that the motion of the proposed bio-soft module has approximate linearity in iso-force and isobaric conditions,and strict linearity in isometric condition.

Optimization and Experimental Study of Bionic Compliant End-effector for Robotic Cherry Tomato Harvesting

High harvesting success rate is part of the key technologies for robotic cherry tomato harvesting,which is closely related to the structural design of the end-effector.To obtain a high success rate of fruit harvesting,this paper presents a compliant end-effector with bio-inspired tarsus compliant gripper inspired by the structure and mechanics of the tarsal chain in the Serica orientalis Motschulsky.Response Surface Methodology(RSM)based on Box Behnken Design(BBD)technique has been used to optimize the key structural parameters of the bionic compliant end-effector for achieving the expected results in pulling pattern for robotic cherry tomato harvesting.Experiments were designed by maintaining three levels of four process parameters—Length of the Offset Segment Tarsomere(OSTL),Angle of the Inclined Segment Tarsomere(ISTA),Thickness of the Extended Segment Tarsomere(ESTT)and Length of the Extended Segment Tarsomere(ESTL).According to the optimization analysis results,the best parameter combination is OSTL 23 mm,ISTA 14°,ESTT 5.0 mm,ESTL 23 mm.Besides,the harvesting performance of the optimized bionic compliant end-effector was verified by experiments.The results indicated the harvesting success rate of fruits with different equatorial diameters was not less than 76%.

Lightweight convolutional neural network models for semanticsegmentation of in-field cotton bolls

Robotic harvesting of cotton bolls will incorporate the benefits of manual picking as well as mechanical harvesting. For robotic harvesting, in-field cotton segmentation with minimal errors is desirable which is a challengingtask. In the present study, three lightweight fully convolutional neural network models were developed for thesemantic segmentation of in-field cotton bolls. Model 1 does not include any residual or skip connections,while model 2 consists of residual connections to tackle the vanishing gradient problem and skip connectionsfor feature concatenation. Model 3 along with residual and skip connections, consists of filters of multiplesizes. The effects of filter size and the dropout rate were studied. All proposed models segment the cotton bollssuccessfully with the cotton-IoU (intersection-over-union) value of above 88.0%. The highest cotton-IoU of91.03% was achieved by model 2. The proposed models achieved F1-score and pixel accuracy values greaterthan 95.0% and 98.0%, respectively. The developed models were compared with existing state-of-the-art networks namely VGG19, ResNet18, EfficientNet-B1, and InceptionV3. Despite having a limited number of trainableparameters, the proposed models achieved mean-IoU (mean intersection-over-union) of 93.84%, 94.15%, and94.65% against the mean-IoU values of 95.39%, 96.54%, 96.40%, and 96.37% obtained using state-of-the-art networks. The segmentation time for the developed models was reduced up to 52.0% compared to state-of-theart networks. The developed lightweight models segmented the in-field cotton bolls comparatively faster andwith greater accuracy. Hence, developed models can be deployed to cotton harvesting robots for real-time recognition of in-field cotton bolls for harvesting.