Weed target detection at seedling stage in paddy fields based on YOLOX.

Weeds are one of the greatest threats to the growth of rice, and the loss of crops is greater in the early stage of rice growth. Traditional large-area spraying cannot selectively spray weeds and can easily cause herbicide waste and environmental pollution. To realize the transformation from large-area spraying to precision spraying in rice fields, it is necessary to quickly and efficiently detect the distribution of weeds. Benefiting from the rapid development of vision technology and deep learning, this study applies a computer vision method based on deep-learning-driven rice field weed target detection. To address the need to identify small dense targets at the rice seedling stage in paddy fields, this study propose a method for weed target detection based on YOLOX, which is composed of a CSPDarknet backbone network, a feature pyramid network (FPN) enhanced feature extraction network and a YOLO Head detector. The CSPDarknet backbone network extracts feature layers with dimensions of 80 pixels ⊆ 80 pixels, 40 pixels ⊆ 40 pixels and 20 pixels ⊆ 20 pixels. The FPN fuses the features from these three scales, and YOLO Head realizes the regression of the object classification and prediction boxes. In performance comparisons of different models, including YOLOv3, YOLOv4-tiny, YOLOv5-s, SSD and several models of the YOLOX series, namely, YOLOX-s, YOLOX-m, YOLOX-nano, and YOLOX-tiny, the results show that the YOLOX-tiny model performs best. The mAP, F1, and recall values from the YOLOX-tiny model are 0.980, 0.95, and 0.983, respectively. Meanwhile, the intermediate variable memory generated during the model calculation of YOLOX-tiny is only 259.62 MB, making it suitable for deployment in intelligent agricultural devices. However, although the YOLOX-tiny model is the best on the dataset in this paper, this is not true in general. The experimental results suggest that the method proposed in this paper can improve the model performance for the small target detection of sheltered weeds and dense weeds at the rice seedling stage in paddy fields. A weed target detection model suitable for embedded computing platforms is obtained by comparing different single-stage target detection models, thereby laying a foundation for the realization of unmanned targeted herbicide spraying performed by agricultural robots.

Innovative DMHS Algorithm Application in Wireless Sensor Networks for Efficient Routing in High-Risk Environments.

Efficient routing is essential for the proper functioning of wireless sensor networks (WSNs). Recent research has focused on optimizing energy and delay for these networks. Nevertheless, there is a dearth of studies that have examined the effects of volatile settings, such as chemical plants, coal mines, nuclear power plants, and battlefields, where connectivity is inconsistent. In such contexts, sensor networks may face security incidents, and environmental factors such as node movement and death can result in dynamic changes to the network topology. A novel design algorithm grounded on Dynamic Minimum Hop Selection (DMHS) was introduced in this paper. The key principle behind DMHS is to use a probabilistic forwarding decision-making process through a distributed route discovery strategy that utilizes dynamically adjusted minimum hop counts of nodes. Simulation results indicate that the life cycle of the DMHS algorithm increases by more than 12% over 700 nodes when compared to the traditional energy-saving algorithm. Furthermore, our algorithm performs better in the average delivery rate of node, and has a 10% to 21% improvement compared to the other algorithms. Overall, the DMHS algorithm represents an important contribution to the development of WSNs that can function robustly in high-risk and unstable environments.

Fully convolutional network for rice seedling and weed image segmentation at the seedling stage in paddy fields.

To reduce the cost of production and the pollution of the environment that is due to the overapplication of herbicide in paddy fields, the location information of rice seedlings and weeds must be detected in site-specific weed management (SSWM). With the development of deep learning, a semantic segmentation method with the SegNet that is based on fully convolutional network (FCN) was proposed. In this paper, RGB color images of seedling rice were captured in paddy field, and ground truth (GT) images were obtained by manually labeled the pixels in the RGB images with three separate categories, namely, rice seedlings, background, and weeds. The class weight coefficients were calculated to solve the problem of the unbalance of the number of the classification category. GT images and RGB images were used for data training and data testing. Eighty percent of the samples were randomly selected as the training dataset and 20% of samples were used as the test dataset. The proposed method was compared with a classical semantic segmentation model, namely, FCN, and U-Net models. The average accuracy rate of the SegNet method was 92.7%, whereas the average accuracy rates of the FCN and U-Net methods were 89.5% and 70.8%, respectively. The proposed SegNet method realized higher classification accuracy and could effectively classify the pixels of rice seedlings, background, and weeds in the paddy field images and acquire the positions of their regions.

A Plant Leaf Geometric Parameter Measurement System Based on the Android Platform.

Automatic and efficient plant leaf geometry parameter measurement offers useful information for plant management. The objective of this study was to develop an efficient and effective leaf geometry parameter measurement system based on the Android phone platform. The Android mobile phone was used to process and measure geometric parameters of the leaf, such as length, width, perimeter, and area. First, initial leaf images were pre-processed by some image algorithms, then distortion calibration was proposed to eliminate image distortion. Next, a method for calculating leaf parameters by using the positive circumscribed rectangle of the leaf as a reference object was proposed to improve the measurement accuracy. The results demonstrated that the test distances from 235 to 260 mm and angles from 0 to 45 degrees had little influence on the leafs' geometric parameters. Both lab and outdoor measurements of leaf parameters showed that the developed method and the standard method were highly correlated. In addition, for the same leaf, the results of different mobile phone measurements were not significantly different. The leaf geometry parameter measurement system based on the Android phone platform used for this study could produce high accuracy measurements for leaf geometry parameters.

Reconstruction of foot and ankle defects with a superthin innervated anterolateral thigh perforator flap.

BACKGROUND: An anterolateral thigh (ALT) perforator flap can be thinned to an extent to which it is vascularised only by the subdermal plexus. This study presents an innervated flap thinning technique and its application for dorsal foot and ankle resurfacing. METHODS: A superthin innervated ALT perforator flap was used to repair the dorsal foot and ankle of 12 patients. The perforators were classified according to their variations in the adipose layer, and the corresponding microdissection technique was then applied. The branch of the lateral femoral cutaneous nerve and its accompanying vessels were adopted to construct a sensory flap. RESULTS: The flap thickness before defatting, which was measured immediately after flap elevation, ranged from 25-45 mm. The average flap thickness after defatting was 4.55 mm (range = 3-6 mm). A total of 11 flaps completely survived, and one flap presented superficial necrosis within a small area (2 cm ×2 cm) in the distal part of the flap. No further flap revision or defatting procedures were required for these patients during an average follow-up period of 16.5 months (range = 10-24 months). In the transferred flap, protective sensibility existed in all cases, and the static two-point discrimination was 13-16 mm. CONCLUSIONS: The superthin innervated ALT perforator flap may be considered as an ideal strategy for foot and ankle reconstruction.

Thinned chimeric radial collateral artery perforator flap in complex distal thumb reconstruction.

INTRODUCTION: Tissue loss accompanied by bone defects in the thumb is a challenging reconstruction problem. Traditional repair methods are unsatisfactory. METHODS: Microsurgical thumb reconstruction was performed using 13 thinned chimeric radial collateral artery (RCA) perforator flaps. The flap was created with a thinned skin paddle and humeral bone segments using independent perforators. Primary defatting was completed when the thickness of the perforator entry was approximately similar to that in the periphery. The posterior cutaneous nerve of the arm was carried to make a sensory flap. Defects were 8.5 × 4.5 cm(2) on average (ranging in size from 4.5 × 1.5 to 15.0 × 6.0 cm(2)), and flap size was 9.0 × 5.5 cm(2) on average (ranging in size from 5.0 × 2.0 to 16.0 × 7.0 cm(2)), whereas the humeral fragments were 2.0 × 1.0 cm(2) on average (ranging in size from 1.5 × 0.5 to 4.0 × 1.5 cm(2)). All data were expressed as mean ± SD. The cosmetic appearance of the donor and recipient sites, Kapandji opposition score and static two-point discrimination of the operated thumb were evaluated during a follow-up visit. RESULTS: Follow-up time was 16.6 months (ranging from 14 to 28 months). Flap thickness before defatting, measured immediately after flap elevation was 14.5 mm (ranging from 10.0 to 25.0 mm). Average flap thickness after defatting was 3.5 mm (ranging from 3.0 to 6.0 mm). Venous congestion occurred in two cases. Successful microsurgical revision was achieved in both cases. All flaps survived. Bone components achieved union in all cases at an average period of 4.8 months (ranging from 3 to 6 months). Based on Kapandji opposition score, the mean thumb opposition score was 6. The mean sensation of flap was 7.5 mm (ranging from 6 to 11 mm). No further flap revision or defatting procedures were required in all cases. Cosmetically acceptable results were achieved for all patients. CONCLUSIONS: Findings proved that thinned chimeric RCA perforator flap is a beneficial microsurgical alternative for reconstructing complex bone and soft tissue defects in thumb.

Modified chimeric radial collateral artery perforator flap for repairing hand composite defects.

BACKGROUND: The authors presented their clinical experience and demonstrated surgical methods for reconstructing complex bone and soft tissue defects of the hand by using modified chimeric radial collateral artery perforator flaps. METHODS: Surgical procedures that employed 16 modified chimeric radial collateral artery perforator flaps and 3 dual paddle flaps were performed in 16 patients. Among the patients, eight had defects in the metacarpal bones and eight had defects in the phalanx bones. The flaps were created with a skin paddle and humeral bone segments by using independent perforators. The flaps ranged in size from 5.5 × 2.0 to 7.5 × 4.5 cm, whereas the humeral fragments ranged in size from 1.5 × 0.5 to 4.0 × 1.5 cm. The pedicle of the flaps was divided and ligated below the level at which the radial collateral artery separates into anterior and posterior branches. The recipient vessels were the proper digital artery, the palmar subcutaneous vein (n = 12), the deep branch of the palmar ulnar artery (n = 4), and the venae comitantes. The cosmetic appearance of both donor and recipient sites was evaluated during a follow-up visit. RESULTS: Postoperative venous congestion occurred in two cases. The venous obstruction was reanastomosed after venous thrombectomy. The procedures were successful in both cases upon examination. All the flaps survived and all the donor sites were closed directly, leaving only a linear scar. Follow-up time ranged from 12 to 28 months. Bone components achieved union in all cases at an average of 5.4 months (ranging from 3 to 6 months). In two cases, the flap was defatted during the late postoperative period. Cosmetically acceptable results were achieved for the rest of the patients. CONCLUSION: The modified chimeric radial collateral artery perforator flap is a good alternative for reconstructing complex bone and soft tissue defects of the hands. LEVEL OF EVIDENCE: This is a level IV, retrospective series.

[Relationship between proximal-tip location and partial necrosis in distally based sural neuro fasciocutaneous flap: an analysis of 157 flaps].

OBJECTIVE: To explore the influence of proximal-tip location on partial necrosis in distally based sural neuro fasciocutaneous flap. METHODS: From April 2001 to May 2009,157 distally based sural neuro fasciocutaneous flaps were conducted to repair the soft tissue defect in distal region of lower leg, ankle and feet in 153 patients. Date of the flaps and the patients were retrospectively analyzed. From the tip of lateral malleolus to the popliteal crease, posterior aspect of the lower leg was equally divided into 9 regions that were 1st to 9th region from inferiorly to superiorly, respectively. The flaps were divided into 2 groups: survival group (including uneventfully survived flaps, flaps with distally epidermal necrosis and with wound dehiscence) and partial necrosis group. Based on the location of the proximal tip of flaps, the flaps were stratified into 4 groups: flaps with the proximal tip locating in the 6th or lower region (group A), the 7th region (group B), the 8th region (group C) and the 9th region (group D). Harvesting the flaps started from exploring the perforator of peroneal vessel in the adipofascial pedicle, then the flaps were elevated retrogradely. RESULTS: Of the 157 flaps, 125 survived uneventfully,8 showed distal epidermal necrosis,wound dehiscence occurred in 6 flaps, 18 flaps (11.5%) showed distal partial necrosis. Partial necrosis occurred in zero of 19 flaps in group A (0), 1 of 44 flaps in group B (2.3% ), 7 of 62 flaps in group C (11.3% ) and 10 of 32 flaps in group D (31.3% ). The differences in partial necrosis rate between group A and group B , group B and group C, were not statistically significant (P > 0.05). Partial necrosis rate was higher in group D than in group C (P = 0.012), it was lower in group A + group B (1.6%) than in group C + group D (18. 1% ) (P = 0. 001). CONCLUSIONS: Distally based sural neuro fasciocutaneous flap can survive reliably when the proximal tip of flap is not beyond the junction between lower 7/9 and upper 2/9 of the lower leg, whereas probability of partial necrosis occurring in the flap increase significantly when the proximal tip of flap locates in upper 1/9 of the lower leg.

[Effect of length-width ratio on partial necrosis in distally based sural neurofasciocutaneous flap: 157 cases].

OBJECTIVE: To explore the effect of length-width ratio (LWR)on partial necrosis in distally based sural neurofasciocutaneous flap. METHODS: Clinical data and operative pictures of 157 distally based sural neurofasciocutaneous flaps were reviewed and analysed. LWR of the flaps ranged from 2.83:1 to 7.14:1. Based on the LWR of the flaps, the flaps were divided into 5 groups: Group A(LWR≤3:1), Group B(3:10.05). Partial necrosis ratio was higher in Group D than that in Group C(P<0.05). It was lower in the LWR<5:1 group (Group A+Group B+Group C) (6.8%, 6/88) than that in the LWR≥5:1 group (Group D+Group E) (17.4%, 12/69) (P<0.05). CONCLUSION: When LWR of the flap is less than 5:1, probability of partial necrosis occurring in distally based sural neurofasciocutaneous flaps is low, and the flap can survive more reliably. The maximum LWR of the surrival flaps can reach 6-7:1.