Determination of mango fruit from binary image using randomized Hough transform

Rizon, M.; Yusri, Nurul Ain Najihah; Kadir, Mohd Fadzil bin Abdul; Mamat, Abd Rasid; Aziz, Azim Zaliha Abd; Nanaa, Kutiba

doi:10.1117/12.2228511

Cited by 11 publications

(4 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The loss function of this algorithm consists of four parts. The first part concerns the prediction of central coordinates, as shown in equation ( 4); the second part concerns about the prediction of the boundary box regression, as shown in equation (5); the third part concerns the prediction of object categories, as shown in equation ( 6); the fourth part concerns the prediction of object confidence, as shown in equation (7).…”

Section: ) Loss Functionmentioning

confidence: 99%

Fast Method of Detecting Tomatoes in a Complex Scene for Picking Robots

Jia

Liu

et al. 2020

IEEE Access

View full text Add to dashboard Cite

At present, there are two main problems with fruit-and vegetable-picking robots. One is that complex scenes (with backlighting, direct sunlight, overlapping fruit and branches, blocking leaves, etc.) obviously interfere with the detection of fruits and vegetables; the other is that an embedded platform needs a lighter detection method due to performance constraints. To address these problems, a fast tomato detection method based on improved YOLOv3-tiny is proposed. First, we improve the precision of the model by improving the backbone network; second, we use image enhancement to improve the detection ability of the algorithm in complex scenes. Finally, we design several groups of comparative experiments to prove the rationality and feasibility of this method. The experimental results show that the f1-score of the tomato recognition model proposed in this paper is 91.92%, which is 12% higher than that of YOLOv3-tiny; the detection speed on a CPU can reach 25 frames/s, and the inferential speed is 40.35 ms, equivalent to that of YOLOv3-tiny. Through comparative experiments, we can see that the comprehensive performance of our method is better than that of other state-of-the-art methods.INDEX TERMS Real-time object detection, deep learning, picking robot, tomato, embedded device.

show abstract

Section: ) Loss Functionmentioning

confidence: 99%

Fast Method of Detecting Tomatoes in a Complex Scene for Picking Robots

Jia

Liu

et al. 2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…24 Rizon et al combined the morphological operator and texture analysis to isolate the overlapped and sheltered mango fruit and then used randomized Hough transform (RHT) to determine the fruit region and the picking point. 25 Luo et al extracted the region of the overlapping grape clusters based on K-means clustering algorithm and separated the region pixels of double overlapping grape clusters based on the contour intersection points and then detected the cutting point of each grape cluster according to the geometric constraint. 26 Fu et al distinguished the fruits calyx from the skin based on color differences and obtained the contact points between the adjacent fruits by analyzing the edge information and then determined the borders of each fruit according to the contact points.…”

Section: Introductionmentioning

confidence: 99%

Location and model reconstruction algorithm for overlapped and sheltered spherical fruits based on geometry

Ouyang

Tian

et al. 2022

International Journal of Advanced Robotic Systems

View full text Add to dashboard Cite

For fruit picking robot, it is an essential prerequisite for achieving fruit picking using machine vision technology to accurately identify the fruits growing in the natural environment. This article presents a vision system of fruit picking robot to perform fruit location and three-dimensional model reconstruction. Firstly, combining the features of color and shape of fruit to reconstruct the actual contour of overlapped and sheltered fruits. Secondly, the least square method was used to reconstruct the three-dimensional model of each fruit according to the spatial coordinates corresponding to image location. Finally, fruit picking experiments in the laboratory environment are used to verify the feasibility of the proposed vision system. Three parameters including Segmentation Error, Intersection Over Union, and False Negative Rate are used to evaluate the performance of the algorithm. The average Segmentation Error, Intersection Over Union, and False Negative Rate of the fruit location algorithm based on geometry were 6.36%, 87.9%, and 6.72%, respectively. The experimental results showed that the average computation time of the algorithm is 3.2 s and the reconstructed three-dimensional model matched the size and position of fruits in the actual scene. The research results can be applied to the vision system of fruit picking robot.

show abstract

“…In particular, the peduncle of grapes is often small and easily obscured by branches and leaves. Therefore, accurate position information relies on extracting the appearance features of fruit, including the color, size, shape, and texture ( Lu and Sang, 2015 ; Rizon et al, 2015 ; Yu et al, 2019 ; Cecotti et al, 2020 ). In the study by Luo et al (2018) , color features were used to extract more effective color components for grapes, which were then segmented to capture images using the k-means clustering algorithm and obtain contours of the grapes.…”

Section: Introductionmentioning

confidence: 99%

Fruit Detection and Pose Estimation for Grape Cluster–Harvesting Robot Using Binocular Imagery Based on Deep Neural Networks

et al. 2021

View full text Add to dashboard Cite

Reliable and robust fruit-detection algorithms in nonstructural environments are essential for the efficient use of harvesting robots. The pose of fruits is crucial to guide robots to approach target fruits for collision-free picking. To achieve accurate picking, this study investigates an approach to detect fruit and estimate its pose. First, the state-of-the-art mask region convolutional neural network (Mask R-CNN) is deployed to segment binocular images to output the mask image of the target fruit. Next, a grape point cloud extracted from the images was filtered and denoised to obtain an accurate grape point cloud. Finally, the accurate grape point cloud was used with the RANSAC algorithm for grape cylinder model fitting, and the axis of the cylinder model was used to estimate the pose of the grape. A dataset was acquired in a vineyard to evaluate the performance of the proposed approach in a nonstructural environment. The fruit detection results of 210 test images show that the average precision, recall, and intersection over union (IOU) are 89.53, 95.33, and 82.00%, respectively. The detection and point cloud segmentation for each grape took approximately 1.7 s. The demonstrated performance of the developed method indicates that it can be applied to grape-harvesting robots.

show abstract

Determination of mango fruit from binary image using randomized Hough transform

Cited by 11 publications

References 1 publication

Fast Method of Detecting Tomatoes in a Complex Scene for Picking Robots

Fast Method of Detecting Tomatoes in a Complex Scene for Picking Robots

Location and model reconstruction algorithm for overlapped and sheltered spherical fruits based on geometry

Fruit Detection and Pose Estimation for Grape Cluster–Harvesting Robot Using Binocular Imagery Based on Deep Neural Networks

Contact Info

Product

Resources

About