Publications
Sort:
Issue
Design and test of quantitative slicing device for pork loin based on contour adaptive shaping
Transactions of the Chinese Society of Agricultural Engineering 2026, 42(14): 156-166
Published: 30 July 2026
Abstract PDF (6.5 MB) Collect
Downloads:4

High precision equipment is often required in meat processing under ever-increasing livestock production at present. Current meat slicing equipment has also suffered from low slicing accuracy and high raw material loss, due to imaging occlusion during the quantitative slicing of irregular boneless meat. In contrast, the imaging after multi-view point cloud (Voint cloud) reconstruction is also plagued by high deployment costs and low adaptability to actual production scenarios. In this study, a high-precision quantitative slicing device was developed for pork loin using contour adaptive shaping. The pork loin was taken as the research object. The low imaging errors and high slicing accuracy were also obtained in the irregular meat products. (1) The device was composed of four components, including a scanning imaging unit, contour adaptive shaping, slicing execution, and a control system. The overall dimensions were defined to determine the key parameters. The effective range of shaping force [Fmin, Fmax] was derived for pork loin using force analysis. A shaping module was integrated with torque, force, and angle sensors to realize the adaptive closed-loop control of shaping parameters. (2) A PLC control system was constructed, including a perception subsystem with various sensors to acquire meat information and shaping motion data, a computing subsystem for image processing and path planning via a computer, and an execution subsystem for the movements of the shaping fixture and slicing mechanism. An adaptive motion of contour shaping and a non-equidistant slicing path planning were designed for the dynamic matching between the cutter and the conveyor belt speed. (3) Pork loin was used as the test sample to verify the actual performance of the shaping module and the slicing device. The conveyor belt and cutter speed matching test showed that the root mean square error (RMSE) of slicing positions ranged from 0.63 to 1.07 mm, and the mean absolute percentage error (MAPE) from 3.98% to 5.15%. There was a linear relationship (v=0.44Me+16.10) between the set slicing weight Me and the conveyor belt speed v, providing for an optimal speed matching scheme for different requirements of quantitative slicing. The verification of shaping effect indicated that the mean relative error (MRE) decreased from 11.13% to 6.24% in the quantitative slicing of pork loin after shaping treatment, whereas the qualification rate (q) increased from 70.12% to 91.43%, indicating an improvement of 4.89% in slicing accuracy and 21.31% in qualification rate, respectively. The optimal slicing effect was achieved in the 60° shaping angle, with the MRE of 6.61% and the q of 89.89%. The continuous slicing tests were conducted to verify the overall performance of the device with the set weights of 100 and 150 g. The slicing accuracy of chilled meat was superior to that of slightly frozen meat. In the quantitative slicing of 100 g chilled meat, the MAPE was 5.57%, and the q was 90.53%. In 150 g chilled meat, the MAPE was 5.11%, and the q was 89.69%. In view of the internal ice crystals and fiber breakage, the slightly frozen meat saw also shared an average increase of 0.76 g in RMSE and 0.84% in MAPE, where the q decreased by an average of 5.06%. The overall processing efficiency of the device reached 134 slices/min, fully meeting the demands of continuous production. In the stability test, the MRE of slicing with set weights of 70, 100, and 150 g ranged from 5.25% to 6.57%, and the coefficient of variation (CV) was all less than 10%, indicating excellent stability of the device. Contour adaptive shaping structure with multiple sensors reduced the slicing errors and raw material loss of conventional slicing equipment in irregular meat processing, thus promoting the transformation of the meat processing industry towards standardization and digitalization. The findings can provide technical references for high-precision quantitative slicing devices for irregular meat products.

Issue
Lightweight instance segmentation method for the key parts of chicken carcass based on improved YOLOv12n-seg
Transactions of the Chinese Society of Agricultural Engineering 2026, 42(9): 217-227
Published: 15 May 2026
Abstract PDF (4.8 MB) Collect
Downloads:1

Accurate segmentation of key parts in the chicken carcass is crucial for intelligent cutting systems in the modern poultry processing industry. However, the false detection, missed detection, and inaccurate segmentation can occur in key parts of chicken carcasses under complex industrial scenarios (e.g., adhesion between wings and drumsticks, occlusion, and uneven lighting). This study aimed to develop a lightweight, high-precision, and real-time instance segmentation model suitable for deployment on intelligent chicken carcass cutting equipment. An enhanced dataset was constructed for chicken carcasses, with emphasis on Sanhuang and white-feathered chickens. Furthermore, the 109 0 original images were expanded into 545 0 images, thus covering three types of scenarios: ambient lighting, carcass occlusion, and compression-induced deformation. Multi-dimensional data augmentation techniques such as geometric transformation, illumination adjustment, and occlusion simulation were adopted to improve the model’s robustness. DEF-YOLO-seg model was developed to improve the YOLOv12n-seg as the baseline: (1) C3k2_DAttention module was designed to fuse the C3k2 module with Deformable Attention (DAttention), which replaced the Area-Attention Enhanced Cross-Feature (A2C2f) module in the lower layer of the backbone network for the feature extraction in the adhered/occluded regions; (2) Efficient Up-Convolution Block (EUCB) was introduced to replace the Upsample module in the neck network, thus reducing computational cost for the feature fusion efficiency; (3) A composite loss function (Focaler-CIoU) with Focaler-IoU and CIoU was constructed for the distribution of easy and difficult samples under complex scenarios. Finally, model training and testing were completed on a hardware platform with an NVIDIA RTX 3090 GPU and an Intel Xeon Platinum 8362 CPU. The DEF-YOLO-seg model achieved a mean Average Precision at an IoU threshold of 0.5 (mAP50) of 95.5% and a mean Average Precision at IoU thresholds from 0.5 to 0.95 (mAP50-95) of 94.1%, which were 1.3 and 2.8 percentage points higher than those of the baseline YOLOv12n-seg, respectively. With a parameter count of 3.3M and a computational complexity of 11GFLOPs, the model’s inference time per image on a local computer was no more than 30 ms. Compared with mainstream models, such as YOLOv9c-seg, YOLOv11n-seg, and YOLOv12n-seg, the improved model maintained lightweight for the superior segmentation accuracy. Furthermore, the parameter sensitivity analysis revealed that the optimal Focaler-CIoU configuration (d=0.22, u=0.73) precisely matched the IoU distribution of chicken carcass data. The task-specific loss function was designed for rather than generic settings. There was the an image-level accuracy of 95.0%. Dice coefficients of the neck, wings, and drumsticks increased from 0.85, 0.83 and 0.78 to 0.93, 0.92 and 0.90, respectively, in the practical production line. The improved model was effectively reduced the missed detection, false detection, and false segmentation of small parts (e.g., neck and shank) under adhesion and occlusion. The DEF-YOLO-seg model also achieved the a better balance among segmentation accuracy, real-time performance, and deployment feasibility in intelligent cutting equipment for chicken carcasses. The findings can provide the technical support to the intelligent upgrading for theof food processing. Future research can focus on cutting path planning, as well as the balance between lightweight and detection accuracy.

Total 2