ABSTRACT
There is an increase in consumption of agricultural produce as a result of the rapidly growing human population, particularly in developing nations. This has triggered high-quality plant phenotyping research to help with the breeding of high-yielding plants that can adapt to our continuously changing climate. Novel, low-cost, fully automated plant phenotyping systems, capable of infield deployment, are required to help identify quantitative plant phenotypes. The identification of quantitative plant phenotypes is a key challenge which relies heavily on the precise segmentation of plant images. Recently, the plant phenotyping community has started to use very deep convolutional neural networks (CNNs) to help tackle this fundamental problem. However, these very deep CNNs rely on some millions of model parameters and generate very large weight matrices, thus making them difficult to deploy infield on low-cost, resource-limited devices. We explore how to compress existing very deep CNNs for plant image segmentation, thus making them easily deployable infield and on mobile devices. In particular, we focus on applying these models to the pixel-wise segmentation of plants into multiple classes including background, a challenging problem in the plant phenotyping community. We combined two approaches (separable convolutions and SVD) to reduce model parameter numbers and weight matrices of these very deep CNN-based models. Using our combined method (separable convolution and SVD) reduced the weight matrix by up to 95% without affecting pixel-wise accuracy. These methods have been evaluated on two public plant datasets and one non-plant dataset to illustrate generality. We have successfully tested our models on a mobile device.
ABSTRACT
Cassava roots are complex structures comprising several distinct types of root. The number and size of the storage roots are two potential phenotypic traits reflecting crop yield and quality. Counting and measuring the size of cassava storage roots are usually done manually, or semi-automatically by first segmenting cassava root images. However, occlusion of both storage and fibrous roots makes the process both time-consuming and error-prone. While Convolutional Neural Nets have shown performance above the state-of-the-art in many image processing and analysis tasks, there are currently a limited number of Convolutional Neural Net-based methods for counting plant features. This is due to the limited availability of data, annotated by expert plant biologists, which represents all possible measurement outcomes. Existing works in this area either learn a direct image-to-count regressor model by regressing to a count value, or perform a count after segmenting the image. We, however, address the problem using a direct image-to-count prediction model. This is made possible by generating synthetic images, using a conditional Generative Adversarial Network (GAN), to provide training data for missing classes. We automatically form cassava storage root masks for any missing classes using existing ground-truth masks, and input them as a condition to our GAN model to generate synthetic root images. We combine the resulting synthetic images with real images to learn a direct image-to-count prediction model capable of counting the number of storage roots in real cassava images taken from a low cost aeroponic growth system. These models are used to develop a system that counts cassava storage roots in real images. Our system first predicts age group ('young' and 'old' roots; pertinent to our image capture regime) in a given image, and then, based on this prediction, selects an appropriate model to predict the number of storage roots. We achieve 91% accuracy on predicting ages of storage roots, and 86% and 71% overall percentage agreement on counting 'old' and 'young' storage roots respectively. Thus we are able to demonstrate that synthetically generated cassava root images can be used to supplement missing root classes, turning the counting problem into a direct image-to-count prediction task.
ABSTRACT
BACKGROUND: Root and tuber crops are becoming more important for their high source of carbohydrates, next to cereals. Despite their commercial impact, there are significant knowledge gaps about the environmental and inherent regulation of storage root (SR) differentiation, due in part to the innate problems of studying storage roots and the lack of a suitable model system for monitoring storage root growth. The research presented here aimed to develop a reliable, low-cost effective system that enables the study of the factors influencing cassava storage root initiation and development. RESULTS: We explored simple, low-cost systems for the study of storage root biology. An aeroponics system described here is ideal for real-time monitoring of storage root development (SRD), and this was further validated using hormone studies. Our aeroponics-based auxin studies revealed that storage root initiation and development are adaptive responses, which are significantly enhanced by the exogenous auxin supply. Field and histological experiments were also conducted to confirm the auxin effect found in the aeroponics system. We also developed a simple digital imaging platform to quantify storage root growth and development traits. Correlation analysis confirmed that image-based estimation can be a surrogate for manual root phenotyping for several key traits. CONCLUSIONS: The aeroponic system developed from this study is an effective tool for examining the root architecture of cassava during early SRD. The aeroponic system also provided novel insights into storage root formation by activating the auxin-dependent proliferation of secondary xylem parenchyma cells to induce the initial root thickening and bulking. The developed system can be of direct benefit to molecular biologists, breeders, and physiologists, allowing them to screen germplasm for root traits that correlate with improved economic traits.
