Core body temperature is one of the most critical physiological indicators of vital signs in livestock and poultry. Effective monitoring and accurate regulation of body temperature can greatly contribute to the health and physiological status of cattle and poultry. The article aims to outline the present research on the various temperature monitoring systems in the sustainable production of livestock and poultry over the last five years. The invasive, wearable and non-contact techniques were also applied to monitor the body temperature of cattle and poultry. Three monitoring systems were evaluated, in terms of data accuracy, easy operation, real-time data, equipment stability, and animal welfare. Three criteria were also set as the technology, cost and application issues. Among them, the invasive monitoring temperature was highly accurate and stable over the long periods, particularly for the scientific research with the high-precision requirements. Nevertheless, this monitoring approach can require the well-trained workers to operate, leading to some pressure on the target animals. Data exchange is also associated with the sensor loss and the displacement issues. Wearable monitoring is easily to operate for the data collection, highly suitable for the long-term continuous temperature monitoring in the large-scale livestock breeding, such as pigs and cattle. However, the monitoring location and usage have the significant impact on the accuracy of temperature monitoring and the comfort of target animals. Non-contact monitoring is extremely sensitive, stress-free and ideal for the grouped monitoring of cattle and poultry animals. The accuracy of temperature measurement is heavily influenced by some factors, such as the external environment, equipment accuracy, testing distance, and testing area. Some recommendations were given for the automatic temperature monitoring in livestock and poultry. In terms of invasive temperature monitoring, some efforts should be focused on the new biocompatible materials, non-invasive/minimally invasive and low-power data communication, in order to reduce the stress response of the target animal for the practicality of overall systems. In terms of wearable temperature monitoring, some emphasis should be put on the optimization of wearing mode, anti-interference algorithms, multi-sensor fusion, and new energy conversion, in order to improve the animal comfort and temperature measurement accuracy. Non-contact temperature monitoring can be required to improve the performance of thermal infrared sensors, with emphasis on the adaptability of temperature measurement for the different species. Deep learning can also be incorporated to construct the more accurate model of temperature monitoring for the high accuracy. The convenient, efficient, low-cost and high-precision devices or application modes can be preferred to fully meet the large-scale needs of temperature measurement in the livestock and poultry breeding. Non-contact temperature monitoring (also known as thermal infrared technology) can be expected to accurately monitor the large-scale livestock and poultry, due to its high sensitivity, absence of stress reaction, and adaptability to group monitoring. The non-contact monitoring can also be standardized to integrate the data collection with the robust algorithm using deep learning. The finding can also provide a strong reference for the more intelligent and efficient temperature monitoring in the livestock and poultry farming.

Recent approaches to the internal quality inspection of apples with the application of hyperspectral imaging technology are highly cost-intensive because of labor involvement for the data collection on a fixed posture and manual selection of the region of interest (RoI). In addition, several studies have repeated the data acquisition for the same apple. Current methods cannot meet the automation requirements of the sorting line. Therefore, this study proposed a novel method for automatically selecting RoI in hyperspectral images of apples with random poses. Firstly, the preliminary RoI selection of apple hyperspectral image was carried out, followed by the performance of histogram statistics of each pixel with spectral intensity at 700 nm wavelength. The top 40% area of the spectral intensity was reserved to obtain the magnitude relationship of the spectral intensity of each pixel point and a morphological erosion operation. Original apple RoI was acquired and overexposed pixels were removed with spectral intensity greater than 3900 (maximum 4095) in the reserved area at 700 nm. Secondly, the relationship between apple size and prediction accuracy was measured for the in-depth RoI analysis. A partial least square regression (PLSR) model was established between the average spectrum and apple sugar content of RoI with different sizes. Finally, the established model with the top 70% of the spectral intensity achieved the best prediction accuracy. Non-destructive estimation of apple sugar content was performed through hyperspectral imaging technology with reference to the proposed RoI selection method. A competitive adaptive reweighted sampling algorithm along the PLSR (CARS-PLSR) model was established after black-and-white correction and standard normal transformation (SNV) preprocessing and obtained the highest prediction accuracy. The determination coefficient of cross-validation (R_cv) and root mean square error of cross-validation (RMSECV) were 0.9595 and 0.3203°Brix, respectively. The determination coefficient of prediction (R_p) was 0.9308, and the root mean square error of prediction (RMSEP) was 0.4681°Brix. Results proved that the auto-selection of RoI is an efficient and accurate method, which can provide a foundation in practical application for online apple grading systems with hyperspectral imaging technology.