Infield Apple Detection and Grading Based on Multi-Feature Fusion

Hu, Guangrui; Zhang, Enyu; Zhou, Jianguo; Zhao, Jing; Gao, Zening; Sugirbay, Adilet; Jin, Hongling; Zhang, Shuo

doi:10.3390/horticulturae7090276

Cited by 25 publications

(13 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In another study, using a deep learning framework and SVM, apples were graded into three categories based on their size, color, shape, and surface flaws with the following detection accuracies for size, color, shape, and surface defects: 99.04%, 97.71%, 98.00%, and 95.85%, respectively. When the feeding interval was shorter than 1.5 s and the walking speed was less than 0.5 m/s, the average grading accuracy was 94.12% …”

Section: Apple Sorting Based On Digital Image Processingmentioning

confidence: 99%

“…Historically, large-scale manual apple sorting has been favored, but it is a painstaking procedure (time-consuming, laborious, and expensive), as well as subjective and inconsistent, all of which degrade the quality of the sorting. Other factors, such as market standards and customer preferences, that demand objective and consistent product classification have urged to investigate nondestructive, accurate, and efficient methods of grading the apples in order to surmount the drawbacks connected with this. ,, …”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

New Insights into Recent Trends and Emerging Technologies in Apple Sorting

Meenu,

Choudhary,

Kaushal

et al. 2024

ACS Food Sci. Technol.

View full text Add to dashboard Cite

Apple sorting is a crucial step for their grading and determination of their commercial value. Sorting techniques are employed at several stages of production process, including harvest, storage, and packing. Manually apples are sorted and graded according to size and color, which is time-consuming, labor-intensive, and error-prone. Therefore, investigating nondestructive, precise, and effective methods of apple sorting is necessary to surmount the drawbacks connected with this. Technology advancements have led to the invention of several automated sorting techniques based on digital image processing, spectral and optical properties that are rapid and efficient compared to manual sorting. These technologies have advanced significantly for assessing the physical and nutritional quality of apples along with internal defects. In the present review, an overview of these advanced apple sorting methods is provided in detail. The contemporary techniques have improved apple sorting accuracy and efficiency while reducing labor costs and improving the quality. Computer vision systems can achieve high levels of accuracy and consistency by enabling nondestructive and quick assessment of many quality metrics, decreasing the subjectivity involved with human sorting. Spectral analysis and hyperspectral imaging have potential for determining interior characteristics like sugar content and ripeness. Robotics, machine learning, and sensor technologies can improve the efficiency, precision, and adaptability of sorting systems by learning from prior sorting data and adjusting to fresh variations in apple quality. In addition, sorting methods combining different techniques need to be developed to detect internal and external quality of apples.

show abstract

Section: Apple Sorting Based On Digital Image Processingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

New Insights into Recent Trends and Emerging Technologies in Apple Sorting

Meenu,

Choudhary,

Kaushal

et al. 2024

ACS Food Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…Although the above research can identify fruits, there are still some drawbacks including a high computational cost, a poor recognition accuracy, a single environmental factor, and a lack of research on multiple clusters of green fruits in complex backgrounds. At present, convolutional neural networks have more mature applications not only in handwritten character recognition ( Soujanya et al., 2022 ; Singh and Chaturvedi, 2023 ) and vehicle detection ( Chen and Li, 2022 ; Gomaa et al., 2022 ), but also in the recognition of fruits such as apples ( Hu et al., 2021 ; Liu et al., 2021 ), pears ( Li et al., 2022 ) and oranges ( Ren and Zhu, 2021 ), but there is no relevant literature on the use of neural networks for green persimmons recognition, it seriously restricts the development of intelligent persimmon orchard operation robots. In this study, a multi-cluster green persimmon recognition algorithm based on improved Faster RCNN was designed by improving the backbone feature extraction network.…”

Section: Introductionmentioning

confidence: 99%

Research on multi-cluster green persimmon detection method based on improved Faster RCNN

et al. 2023

View full text Add to dashboard Cite

To address the problem of accurate recognition and localization of multiple clusters of green persimmons with similar color to the background under natural environment, this study proposes a multi-cluster green persimmon identification method based on improved Faster RCNN was proposed by using the self-built green persimmon dataset. The feature extractor DetNet is used as the backbone feature extraction network, and the model detection attention is focused on the target object itself by adding the weighted ECA channel attention mechanism to the three effective feature layers in the backbone, and the detection accuracy of the algorithm is improved. By maximizing the pooling of the lower layer features with the added attention mechanism, the high and low dimensions and magnitudes are made the same. The processed feature layers are combined with multi-scale features using a serial layer-hopping connection structure to enhance the robustness of feature information, effectively copes with the problem of target detection of objects with obscured near scenery in complex environments and accelerates the detection speed through feature complementarity between different feature layers. In this study, the K-means clustering algorithm is used to group and anchor the bounding boxes so that they converge to the actual bounding boxes, The average mean accuracy (mAP) of the improved Faster RCNN model reaches 98.4%, which was 11.8% higher than that of traditional Faster RCNN model, which also increases the accuracy of object detection during regression prediction. and the average detection time of a single image is improved by 0.54s. The algorithm is significantly improved in terms of accuracy and speed, which provides a basis for green fruit growth state monitoring and intelligent yield estimation in real scenarios.

show abstract

“…The overall mAP was less than 74%. Guangrui Hu et al [16] used the TensorFlow deep learning framework and SSD deep learning algorithm to identify apple surface defects. Yanfei Li et al [17] proposed a fast classification model of apple quality based on a convolutional neural network (CNN) and compared it with the Google InceptionV3 model and HOG/GLCM + SVM.…”

Section: Introductionmentioning

confidence: 99%

Real-Time Grading of Defect Apples Using Semantic Segmentation Combination with a Pruned YOLO V4 Network

Liang

Jia

Huang

et al. 2022

Foods

View full text Add to dashboard Cite

At present, the apple grading system usually conveys apples by a belt or rollers. This usually leads to low hardness or expensive fruits being bruised, resulting in economic losses. In order to realize real-time detection and classification of high-quality apples, separate fruit trays were designed to convey apples and used to prevent apples from being bruised during image acquisition. A semantic segmentation method based on the BiSeNet V2 deep learning network was proposed to segment the defective parts of defective apples. BiSeNet V2 for apple defect detection obtained a slightly better result in MPA with a value of 99.66%, which was 0.14 and 0.19 percentage points higher than DAnet and Unet, respectively. A model pruning method was used to optimize the structure of the YOLO V4 network. The detection accuracy of defect regions in apple images was further improved by the pruned YOLO V4 network. Then, a surface mapping method between the defect area in apple images and the actual defect area was proposed to accurately calculate the defect area. Finally, apples on separate fruit trays were sorted according to the number and area of defects in the apple images. The experimental results showed that the average accuracy of apple classification was 92.42%, and the F1 score was 94.31. In commercial separate fruit tray grading and sorting machines, it has great application potential.

show abstract

Infield Apple Detection and Grading Based on Multi-Feature Fusion

Cited by 25 publications

References 31 publications

New Insights into Recent Trends and Emerging Technologies in Apple Sorting

New Insights into Recent Trends and Emerging Technologies in Apple Sorting

Research on multi-cluster green persimmon detection method based on improved Faster RCNN

Real-Time Grading of Defect Apples Using Semantic Segmentation Combination with a Pruned YOLO V4 Network

Contact Info

Product

Resources

About