Pre-harvest fruit yield estimation is useful to guide harvesting and marketing resourcing, but machine vision estimates based on a single view from each side of the tree (“dual-view”) underestimates the fruit yield as fruit can be hidden from view. A method is proposed involving deep learning, Kalman filter, and Hungarian algorithm for on-tree mango fruit detection, tracking, and counting from 10 frame-per-second videos captured of trees from a platform moving along the inter row at 5 km/h. The deep learning based mango fruit detection algorithm, MangoYOLO, was used to detect fruit in each frame. The Hungarian algorithm was used to correlate fruit between neighbouring frames, with the improvement of enabling multiple-to-one assignment. The Kalman filter was used to predict the position of fruit in following frames, to avoid multiple counts of a single fruit that is obscured or otherwise not detected with a frame series. A “borrow” concept was added to the Kalman filter to predict fruit position when its precise prediction model was absent, by borrowing the horizontal and vertical speed from neighbouring fruit. By comparison with human count for a video with 110 frames and 192 (human count) fruit, the method produced 9.9% double counts and 7.3% missing count errors, resulting in around 2.6% over count. In another test, a video (of 1162 frames, with 42 images centred on the tree trunk) was acquired of both sides of a row of 21 trees, for which the harvest fruit count was 3286 (i.e., average of 156 fruit/tree). The trees had thick canopies, such that the proportion of fruit hidden from view from any given perspective was high. The proposed method recorded 2050 fruit (62% of harvest) with a bias corrected Root Mean Square Error (RMSE) = 18.0 fruit/tree while the dual-view image method (also using MangoYOLO) recorded 1322 fruit (40%) with a bias corrected RMSE = 21.7 fruit/tree. The video tracking system is recommended over the dual-view imaging system for mango orchard fruit count.

Eight depth cameras varying in operational principle (stereoscopy: ZED, ZED2, OAK-D; IR active stereoscopy: Real Sense D435; time of flight (ToF): Real Sense L515, Kinect v2, Blaze 101, Azure Kinect) were compared in context of use for in-orchard fruit localization and sizing. For this application, a specification on bias-corrected root mean square error of 20 mm for a camera-to-fruit distance of 2 m and operation under sunlit field conditions was set. The ToF cameras achieved the measurement specification, with a recommendation for use of Blaze 101 or Azure Kinect made in terms of operation in sunlight and in orchard conditions. For a camera-to-fruit distance of 1.5 m in sunlight, the Azure Kinect measurement achieved an RMSE of 6 mm, a bias of 17 mm, an SD of 2 mm and a fill rate of 100% for depth values of a central 50 × 50 pixels group. To enable inter-study comparisons, it is recommended that future assessments of depth cameras for this application should include estimation of a bias-corrected RMSE and estimation of bias on estimated camera-to-fruit distances at 50 cm intervals to 3 m, under both artificial light and sunlight, with characterization of image distortion and estimation of fill rate.

Forward estimates of harvest load require information on fruit size as well as number. The task of sizing fruit and vegetables has been automated in the packhouse, progressing from mechanical methods to machine vision over the last three decades. This shift is now occurring for size assessment of fruit on trees, i.e., in the orchard. This review focuses on: (i) allometric relationships between fruit weight and lineal dimensions; (ii) measurement of fruit lineal dimensions with traditional tools; (iii) measurement of fruit lineal dimensions with machine vision, with attention to the issues of depth measurement and recognition of occluded fruit; (iv) sampling strategies; and (v) forward prediction of fruit size (at harvest). Commercially available capability for in-orchard fruit sizing is summarized, and further developments of in-orchard fruit sizing by machine vision are anticipated.

Malaria in the rural and remote regions of tropical countries remain a major public health challenge. Early diagnosis and prompt effective treatment are the basis for the management of malaria and for reducing malaria mortality and morbidity worldwide and the key to malaria elimination. While Rapid Diagnostic Test (RDT) remains the current mainstay testing malaria infections, it is usually used in conjunction with clinical findings and lab tests of blood films through Microscopy- the gold standard of malaria diagnosis. Recent reports suggest that the accuracy of RDTs could be compromised due to parasite antigen gene deletion(s), and the lack of expertise and high turnover time makes microscopy impractical to be used in rural and remote areas which impede the diagnosis and treatment of the disease. Delay in receiving treatment for uncomplicated malaria is reported to increase the risk of developing severe malaria and mortality. Thus, the need to develop advanced, faster, and smarter tools for malaria diagnosis is paramount, specially to reinforce the gold standard method, i.e., malaria microscopy which is a full-proof tool given the limitations be addressed. Deep learning-based methods have proven to provide human expert level performance on object detection/classification on image data. Such methods can be utilized for automation of repetitive task in assessing large number of microscope images of blood samples. In this paper, we propose a novel approach to improve the performance of deep learning models through consistent labelling of ground truth bounding box for the task of pathogen detection on microscope images of thick blood smears. Recommendations are made on the reliability and repeatability testing of the trained models. A custom deep learning architecture (YOLO-mp) is developed based on the design criteria of optimizing accuracy and speed of detection with minimal resources. The custom three-layered YOLO-mp-3l and four-layered YOLO-mp-4l models achieved the best mAP scores of 93.99 (@IoU=0.5) and 94.07 (@IoU=0.5), respectively outperforming standard YOLOv4 (mAP 92.56 @IoU=0.5) for detection of malaria pathogen on a public dataset of thick blood smear microscope images captured using phone camera. YOLO-mp-3l (BFLOPs =21.8, model size =24.5Mb) and YOLO-mp-4l (BFLOPs=24.477, model size =25.4Mb) outperformed standard YOLOv4 (BFLOPs=127.232, model size = 244Mb) in terms of computation and memory requirements proving them suitable to run on low resource devices.