Thermal imaging and computer vision technologies for the enhancement of pig husbandry: a review

Md Nasim Reza,Md Razob Ali,Samsuzzaman,Md Shaha Nur Kabir,Md Rejaul Karim,Shahriar Ahmed,Hyunjin Kyoung,김국환,Sun-Ok Chung 한국축산학회 2024 한국축산학회지 Vol.66 No.1

Pig farming, a vital industry, necessitates proactive measures for early disease detection and crush symptom monitoring to ensure optimum pig health and safety. This review explores advanced thermal sensing technologies and computer vision-based thermal imaging techniques employed for pig disease and piglet crush symptom monitoring on pig farms. Infrared thermography (IRT) is a non-invasive and efficient technology for measuring pig body temperature, providing advantages such as non-destructive, long-distance, and high-sensitivity measurements. Unlike traditional methods, IRT offers a quick and labor-saving approach to acquiring physiological data impacted by environmental temperature, crucial for understanding pig body physiology and metabolism. IRT aids in early disease detection, respiratory health monitoring, and evaluating vaccination effectiveness. Challenges include body surface emissivity variations affecting measurement accuracy. Thermal imaging and deep learning algorithms are used for pig behavior recognition, with the dorsal plane effective for stress detection. Remote health monitoring through thermal imaging, deep learning, and wearable devices facilitates non-invasive assessment of pig health, minimizing medication use. Integration of advanced sensors, thermal imaging, and deep learning shows potential for disease detection and improvement in pig farming, but challenges and ethical considerations must be addressed for successful implementation. This review summarizes the state-of-the-art technologies used in the pig farming industry, including computer vision algorithms such as object detection, image segmentation, and deep learning techniques. It also discusses the benefits and limitations of IRT technology, providing an overview of the current research field. This study provides valuable insights for researchers and farmers regarding IRT application in pig production, highlighting notable approaches and the latest research findings in this field.

3D LiDAR-based Quantification of Phenotypic Traits and Land Characteristics in Rice Farming

( Md Rejaul Karim ),( Mohammod Ali ),( Shahriar Ahmed ),( Md Shaha Nur Kabir ),( Md Nasim Reza ),( Justin Sung ),( Sun-ok Chung ) 한국농업기계학회 2023 한국농업기계학회 학술발표논문집 Vol.28 No.2

Phenotypic and land characteristics information plays a crucial role in effective management of rice farming. The utilization of LiDAR based object recognition as well as visualization provides a rapid and precise assessment of the phenotypic traits of rice plants. This study aimed to quantify the rice plant phenotypic and land characteristics using a 3D LiDAR. A data collection structure made of aluminum profile and a LiDAR sensor (i.e., VLP-16) mounted on the structure was used to collect 3D point cloud data from rice field. A rice field of RDA at Iksan in Korea was selected for data acquisition. Ten numbers of small plots considering the area of LiDAR data frames exhibiting diverse plant height, shapes, and sizes were randomly selected. From each LiDAR scanned data frame, a region of interest (RoI) segmented for sensor based processing and measurements. Commercial software utilized for segmentation and python-based programing codes also applied to process the collected data for visualization and measurements. The accuracy of the estimated outputs from the point cloud was evaluated by comparison with measured values collected randomly from ten spots remaining in the sensor-based data frame. The estimated plant heights from the point cloud were 0.84±0.03 m, while the measured heights were 0.77±0.03 m. The root mean square error (RMSE) for plant height estimation was 0.08 m, and the simple linear coefficient of determination (r²) was 0.88. Regarding the segment wise canopy volume, point cloud estimations were 1.01±0.06 m³, compared to the measured volume of 1.18±0.03 m³. The RMSE for canopy volume estimation was 0.18 m³, with r² of 0.87, indicating a high level of accuracy. For hill-to-hill distance and intra-row spacing, the point cloud measurements were 0.35±0.01 m, and 0.34±0.02 m, respectively, while the measurements were 0.30±0.03 m, and 0.30±0.03 m, respectively. The RMSE and r² for hill to hill distance were 0.04 m and 0.92, respectively, and for row distance, 0.03 m and 0.87, respectively. Despite minor differences, there was a strong relationship and close agreement between the estimation using point cloud data and measurements. The findings highlight the reliability and efficiency of the 3D LiDAR technology for accurately measuring phenotypic traits and land characteristics for maximizing rice cultivation.

Automatic Counting of Rice Plant Numbers After Transplanting Using Low Altitude UAV Images

Md Nasim Reza,In Seop Na,Kyeong-Hwan Lee 한국콘텐츠학회 2017 International Journal of Contents Vol.13 No.3

AI-Enabled Real-Time Pig Disease Detection and Management

( Md Nasim Reza ),( Sumaiya Islam ),( Md Razob Ali ),( Samsuzzaman ),( Md Shaha Nur Kabir ),( Minho Song ),( Gookhwan Kim ),( Sun-ok Chung ) 한국농업기계학회 2023 한국농업기계학회 학술발표논문집 Vol.28 No.2

Surveillance cameras are becoming crucial tools for early livestock disease detection, offering the potential to reduce the negative impact on animal health and the economy in livestock production. This study focused on detecting pig disease symptoms, serving as an initial exploration for practical implementation on pig farms. The aim was to develop an AI-based approach using various video and acoustic sensors in real farm environments. The setup includes two RGB cameras for top and side views, a thermal sensor, and a sound sensor, all controlled by a microcontroller. The collected audio, video, and temperature data are processed in real-time. Using RGB and infrared camera feeds, along with audio analysis, we developed a system to recognize pigs and identify illness states in the video stream. We employed a single-shot multibox (SSD) architecture with MobileNet V2 for video stream processing, achieving an accuracy of 93.6% for pig recognition. The system demonstrated an 89.6% mean average accuracy (mAP) with a frame rate of 21 for disease detection. When tested on sound data, it achieved an average F1-score of 83.7%, with recognition accuracies of 67.5% for snoozing, 74.8% for coughs, 72.9% for crushing sounds, and 82.3% for screaming. Detection accuracy was affected by blurry video and background noises. This research advances precision livestock farming for pig health and disease prevention.

Analysis of Rice Grain and Morphological Characteristics in 3D Model Generated from Multi-Camera UAV Images

( Md Nasim Reza ),( Sunwook Baek ),( Sang-eon Oh ),( Kyeonghwan Lee ) 한국농업기계학회 2019 한국농업기계학회 학술발표논문집 Vol.24 No.1

Advanced crop management is an important issue in the field of precision agriculture. Rice yield estimation during crop growing season is a significant indicator and plant parameters, such as leaf area index, height, stem diameter, biomass, etc. are highly related with yield to analyze the impact of crop management practices. The traditional approaches to attain these parameters from rice field are very challenging, laborious and time-consuming. 3D geometric information based on UAV imagery of rice plant can be a relevant solution to estimate these valuable parameters. The objective of this study was to evaluate the application of the 3D model generated from low altitude UAV based multi-camera images to analyze rice grain and estimate plant height and biomass. We flied DJI S1000 octocopter with a camera bracket beneath it. The bracket was designed to hold three Sony alpha a5100 digital cameras with one nadir and two oblique views on the ground. RGB images were taken from a height of 10m using autoflight and predesigned flight path. With these images being processed into 3D point clouds, which were subsequently used to generate 3D model of rice plants. We designed an algorithm to recognize grain and measure height and biomass from the 3D model. The classification of ground and vegetation part was done to compute vegetation height. To identify rice grain in panicles, we applied regionprops and object featured segmentation. Then, we applied 3D surface plotting and voxelization to measure biomass volume. The proposed method showed a strong correlation between observed and actual measurement of rice grain count, height and biomass. Multi-camera imaging was shown to be effective at 3D modelling and estimating morphological characteristics of the rice plant.

Lab Color Space based Rice Yield Prediction using Low Altitude UAV Field Image

( Md Nasim Reza ),( Inseop Na ),( Sunwook Baek ),( In Lee ),( Kyeonghwan Lee ) 한국농업기계학회 2017 한국농업기계학회 학술발표논문집 Vol.22 No.1

Prediction of rice yield during a growing season would be very helpful to magnify rice yield as it also allows better farm practices to maximize yield with greater profit and lesser costs. UAV imagery based automatic detection of rice can be a relevant solution for early prediction of yield. So, we propose an image processing technique to predict rice yield using low altitude UAV images. We proposed L*a*b* color space based image segmentation algorithm. All images were captured using UAV mounted RGB camera. The proposed algorithm was developed to find out rice grain area from the image background. We took RGB image and applied filter to remove noise and converted RGB image to L*a*b* color space. All color information contain in both a* and b* layers and by using k-mean clustering classification of these colors were executed. Variation between two colors can be measured and labelling of pixels was completed by cluster index. Image was finally segmented using color. The proposed method showed that rice grain could be segmented and we can recognize rice grains from the UAV images. We can analyze grain areas and by estimating area and volume we could predict rice yield.

Automatic Counting of Rice Plant Numbers After Transplanting Using Low Altitude UAV Images

Reza, Md Nasim,Na, In Seop,Lee, Kyeong-Hwan The Korea Contents Association 2017 International Journal of Contents Vol.13 No.3

Rice plant numbers and density are key factors for yield and quality of rice grains. Precise and properly estimated rice plant numbers and density can assure high yield from rice fields. The main objective of this study was to automatically detect and count rice plants using images of usual field condition from an unmanned aerial vehicle (UAV). We proposed an automatic image processing method based on morphological operation and boundaries of the connected component to count rice plant numbers after transplanting. We converted RGB images to binary images and applied adaptive median filter to remove distortion and noises. Then we applied a morphological operation to the binary image and draw boundaries to the connected component to count rice plants using those images. The result reveals the algorithm can conduct a performance of 89% by the F-measure, corresponding to a Precision of 87% and a Recall of 91%. The best fit image gives a performance of 93% by the F-measure, corresponding to a Precision of 91% and a Recall of 96%. Comparison between the numbers of rice plants detected and counted by the naked eye and the numbers of rice plants found by the proposed method provided viable and acceptable results. The $R^2$ value was approximately 0.893.

The Analysis of Rice Transplant Characteristics by using Low Altitude UAV Images

( Md Nasim Reza ),( Inseop Na ),( Sunwook Baek ),( Kyeonghwan Lee ) 한국농업기계학회 2016 한국농업기계학회 학술발표논문집 Vol.21 No.2

Manual field based monitoring is labor intensive, expansive and time consuming. To overcome this situation automatic monitoring of rice seedlings lane and growth can be done. So, we proposed an image processing technique to detect and count the rice plant lane using low altitude Unmanned Aerial vehicle (UAV) images. The main objective of the study is to make an image processing technique based on horizontal and vertical projection of low altitude RGB images obtained from UAV for automatic rice plant lane detection. The algorithm was developed as follows: the initial RGB images, convert the images to gray or binary, noise filter, vertical and horizontal projection, detection of plant lane and calculate the length width, removal of false lane, final result. An adaptive median filter was used to remove the noise and image projection method was applied to optimize the plant lane. The result of the image projection was used to detect the plant lanes in the field. The accuracy of the result was compared with the ground truth. Our method showed that it is efficient for detecting and counting the rice plant lanes. The proposed method have shown that it is able to detect and count lanes without any mechanical interferences and it may be used as an automated tool for different crops.

Rice yield estimation based on K-means clustering with graph-cut segmentation using low-altitude UAV images

Reza, Md Nasim,Na, In Seop,Baek, Sun Wook,Lee, Kyeong-Hwan Elsevier 2019 Biosystems engineering Vol.177 No.-

<P><B>Abstract</B></P> <P>Predicting the harvest yield enables farm practices to be modified throughout the growing season, with potential to increase the final yield. Unmanned aerial vehicle (UAV) based remote sensing is a promising way to estimate crop yields. In this study, rice yield was estimated by segmenting grain areas using low altitude RGB images collected using a rotary-wing type UAV. In particular, an image processing method that combines K-means clustering with a graph-cut (KCG) algorithm was proposed to segment the rice grain areas. The graph-cut algorithm was applied to extract the foreground and background of the images. The foreground RGB images were converted to the Lab colour space and then K-means clustering was used to label pixels based on colour information. The area of the rice grains in the images was calculated from the clustered images. Using this grain area information, the rice yield of the field could be estimated. Experiments show that the proposed method can segment the grain areas with a relative error of 6%–33%, and it improved the relative error of the previous method (by 1%–31%). The coefficient of determination between the results of the proposed method and the ground truth was found to be 0.98. Furthermore, the relative error of the yield estimation for four field sections was 21%–31%. The results indicate that the UAV image-based grain segmentation has the potential to estimate rice yield accurately and conveniently.</P> <P><B>Highlights</B></P> <P> <UL> <LI> Rice yield estimation based on segmented rice grain area using UAV images. </LI> <LI> Improve the accuracy of segmented rice grain area by using estimated foreground objects. </LI> <LI> Estimate the rice yield during whole stage of growing season. </LI> <LI> Measure the optimal estimation stage for rice yield in the lifecycle of rice plant. </LI> </UL> </P>

Construction and basic performance test of an ICT-based irrigation monitoring system for rice cultivation in UAE desert soil

ALI MOHAMMOD,Reza Md Nasim,Kiraga Shafik,ISLAM MD NAFIUL,Chowdhury Milon,정재혁,정선옥 충남대학교 농업과학연구소 2021 Korean Journal of Agricultural Science Vol.48 No.4

An irrigation monitoring system is an efficient approach to save water and to provide effective irrigation scheduling for rice cultivation in desert soils. This research aimed to design, fabricate, and evaluate the basic performance of an irrigation monitoring system based on information and communication technology (ICT) for rice cultivation under drip and micro-sprinkler irrigation in desert soils using a Raspberry Pi. A data acquisition system was installed and tested inside a rice cultivating net house at the United Arab Emirates University, Al-Foah, AlAin. The Raspberry Pi operating system was used to control the irrigation and to monitor the soil water content, ambient temperature, humidity, and light intensity inside the net house. Soil water content sensors were placed in the desert soil at depths of 10, 20, 30, 40, and 50 cm. A sensor-based automatic irrigation logic circuit was used to control the actuators and to manage the crop irrigation operations depending on the soil water content requirements. A developed webserver was used to store the sensor data and update the actuator status by communicating via the Pi-embedded Wi-Fi network. The maximum and minimum average soil water contents, ambient temperatures, humidity levels, and light intensity values were monitored as 33.91 ± 2 to 26.95 ± 1%, 45 ± 3 to 24 ± 3℃, 58 ± 2 to 50 ± 4%, and 7160 - 90 lx, respectively, during the experimental period. The ICT-based monitoring system ensured precise irrigation scheduling and better performance to provide an adequate water supply and information about the ambient environment.