The current recommendations for nitrogen (N) doses may not contemplate the complex dynamics of N in the soil, what may imply in insufficiency to the crop demand. This study aimed to evaluate the use of two portable chlorophyll meters (Minolta SPAD 502 and ClorofiLOG model CFL 1030), after defining the nitrogen sufficiency index (NSI), to estimate the nitrogen fertilization factor in modern bean genotypes. A block design was used, with four replications. Each experiment consisted of four treatments, with doses of 0, 5, 10 and 20 kg ha-1 of N for each 1 % considered below the NSI defined as adequate for each cultivar. At the end of the cycle, the production components and yield were evaluated. It was not possible to obtain precise adjustments in the methodology for the TAA Gol cultivar. The fertilization factor of 12.5 kg ha-1 of N for each 1 % below the NSI of 90 % proved to be efficient for the Pérola cultivar, when the SPAD-502 chlorophyll meter was used. In genotypes with size and cycle characteristics similar to those of the BRSMG Uai cultivar and the VR 20 strain, using a fertilization factor between 14 and 16 kg ha-1 of N, the NSI of 95 % must be reduced, in order to increase the efficiency of the nitrogen fertilization.

Portable chlorophyll meters can be used to improve the prediction of nitrogen (N) doses for common bean. This study aimed to evaluate two chlorophyll meters (Minolta SPAD-502 and ClorofiLOG CFL 1030) to predict topdressing N doses for the Pérola, TAA Gol and BRSMG Uai cultivars and for the VR 20 line. Eight field experiments (four genotypes and two devices) were carried out in a randomized blocks design, with four replicates, being the treatments topdressing N applications corresponding to four nitrogen sufficiency indices. There were discrepancies between the indices obtained by the different devices for the same genotype, as well as among those obtained with the same device for the different genotypes studied. Thus, when the index is used to define the N doses, the genotype and the chlorophyll meter used must be considered. In addition, the recommended index of 95 % should not be generalized to all the cultivars and should be reduced.

The present works brings the development of a perception system for crop detection, activation of a single nozzle and simulation for a pesticide sprayer. The purpose is to embed the system on the agricultural machine to avoid the waste of pesticide by applying it only where it is necessary, being less aggressive with the soil. Our motivation is to implement simply advanced technique capable of being adopted by family farmers, with low cost and high efficiency. There is a hardware cost limitation, such as the use of a single camera to the proposal's motivation of being viable for family farmers. Its implementation occurs by using simple hardware that runs computer vision software that applies crop row detection techniques and triggers logic to the nozzles. The results showed that the proposal can identify the crops and activate each nozzle individually. However, the processing time test shows that the algorithm needs an implementation improvement.