The choice of the radius parameters proves to be dependent on the application.While the normal radius rn can usually simply be set to a value equal or greater than the resolution, the support radius r has to be adjusted to the special requirements. If the plant organs that are to be distinguished vary only on a small scale, like on the flower data set, this has to be reflected by a smaller choice of r.As expected, choosing an SVM as more sophisticated classification method makes the choice of the descriptor almost irrelevant, as all of them achieve very good results. But for an SVM, a gold standard has to be prepared and depending on the use case, this can be hard . Fortunately, even in the case of unsupervised classification FPFHs yield very good results.The evaluation on the representative sets shows a clear ranking for the SVM-based classification: FPFHs perform best, while the other descriptors all yield results similar to each other. In the case of the k-means clustering, we have on average the following ranking: FPFHs > PFHs > Spin Images > RoPS, SHOT.There are slight deviations, e.g., the Spin Images show the worst results of all descriptors on the Leaves data set, but reach almost the same quality of results as FPFHs on the Branch data set.

The same effect can be seen in the SVM results.This suggests that the resolution chosen for the Spin Images in this paper is better suited to distinguish between round and flat or cylindrical shapes than between flat and cylindrical shapes only.All in all and despite the exemplary character of the evaluation the results clearly suggest using FPFHs as descriptor of choice when compared with SHOT,RoPS and Spin Images.In applications like scan registration, both RoPS and SHOT descriptor were found to outperform PFHs and FPFHs. In contrast to that, we strive to classify the whole set of points and assign them to the corresponding plant organ.This means that the descriptor has to be able to generalize over different sizes of plant organs. Additionally, scans can not be expected to be perfect, as they have to be taken in the field and for a high number of plants. Parts of the plant can be occluded by other parts and holes in the data are possible. The descriptor has to be robust against these issues. The good performance of the FPFHs and PFHs in our application together with the worse performance of RoPS and SHOT descriptors therefore hints that FPFHs and PFHs seem to have the generality that makes them less suitable for applications where different points on a similar shaped surface have to be distinguished, but optimal for point classification in the context of precision farming.

Almost all of the pig can be used as food. There are many styles of farming: intensive commercial units, commercial free range enterprises. Although all these forms are in use today, intensive farms are the most popular due to their abilities to raise a large amount of pigs in a very cost-efficient manner. For example, only3% of UK pigs spend their entire lives outdoors .There are challenges facing in intensive pig farming. For instance, hogs in intensive farms tendto produce 23.5 piglets per year. Sow death rates have nearly doubled from 5.8%- 10.2% from 2013 to 2016 . Researchers and veterinarians are seeking ways with genetic manipulation positively impact the health of the hogs and benefit the hog business .China has the world’s largest herd and has been increasing its pig imports during its economic development. Chinese pig farming industry is growing and rapidly shifting to modernization through introducing the lean production concept and embracing the ICT .There are many efforts made for tackling the farming challenges for the increasing competition, farming cost, and guaranteeing hog health.The largest exporters of pigs are the United States, the European Union, and Canada.China’s pig farming is relatively limited in the use of new technology for a long time.

The State Council of the People’s Republic of China issued the suggestions on stabilizing pig farming and promoting its transformation and upgrading with the use of ICT, artificial intelligence and automation, and proposing to accelerate the modernization of pig farming systems in 2019 .A modern smart pig farming system should provide services for genetic manipulation, frozen sperm usage, mass production, precise feeding , remote diagnosis and treatment , daily weight gaining and cost control , performance management, talent training, smart pig farm construction,etc. For an advanced smart pig farming system, we realize that it’s better to make use of digitalization, infomatization and Internet of things to collect data on piglife cycle for pig industry. With those data to guide pig farming, a pig farm could achieve lean management with high production efficiency and quality.Realizing the challenges faced by pig farming and the opportunities given by the next generation artificial intelligence, we design an Industrial Internet Platform for Massive Pig Farming . The implementation verifies that the application of new technology such as the Industrial Internet really promotes pigfarming industry.