This work introduces a new vision-based approach for estimating chlorophyll contents in a plant leaf using reflectance and transmittance as base parameters. Images of the top and underside of the leaf are captured. To estimate the base parameters (reflectance/transmittance), a novel optical arrangement is proposed. The chlorophyll content is then estimated by using linear regression where the inputs are the reflectance and transmittance of the leaf. Performance of the proposed method for chlorophyll content estimation was compared with a spectrophotometer and a Soil Plant Analysis Development (SPAD) meter. Chlorophyll content estimation was realized for Lactuca sativa L., Azadirachta indica, Canavalia ensiforme, and Lycopersicon esculentum. Experimental results showed that—in terms of accuracy and processing speed—the proposed algorithm outperformed many of the previous vision-based approach methods that have used SPAD as a reference device. On the other hand, the accuracy reached is 91% for crops such as Azadirachta indica, where the chlorophyll value was obtained using the spectrophotometer. Additionally, it was possible to achieve an estimation of the chlorophyll content in the leaf every 200 ms with a low-cost camera and a simple optical arrangement. This non-destructive method increased accuracy in the chlorophyll content estimation by using an optical arrangement that yielded both the reflectance and transmittance information, while the required hardware is cheap.

No abstract

ResumenLa presente inves ti ga ción sugiere una contri bu ción en la apli ca ción de modelos de pronós ticos. El modelo propuesto se desa rrolla con el propó sito de ajustar la proyección de la demanda al esce nario de las empresas y se funda menta en tres consi de raciones que provocan que en muchos casos los pronós ticos de la demanda disten de la realidad, como son: 1) uno de los problemas más difí ciles de modelar en los pronósticos es la incer ti dumbre rela cio nada con la infor ma ción dispo nible; 2) los métodos tradi cio nal mente utili zados por las empresas, para la proyec ción de la demanda, se basan prin ci pal mente en el compor ta miento pasado del mercado (demanda histórica), y 3) estos métodos no consi deran en su análisis a los factores que están influyendo para que se dé el compor ta miento obser vado. Por lo tanto, el modelo propuesto se basa en la imple men ta ción de lógica difusa, inte gran do las prin ci pales varia bles que afectan el compor ta miento de la demanda del mercado y que no son consi de radas en los métodos esta dís ticos clásicos. El modelo se aplicó a una embo tella dora de bebidas carbo na tadas y con el ajuste de la proyec ción de la demanda se obtuvo un pronós tico más confiable.Descrip tores: planea ción de la produc ción, pronós tico agre gado, modelo de pronós tico, demanda, lógica difusa e inte li gencia compu ta cional en inge niería indus trial. inge nie ría INVESTIGACIÓN Y TECNOLOGÍAAbstract This re search sug gests a con tri bu tion in the im ple men ta tion of fore cast ing mod els. The proposed model is de vel oped with the aim to fit the pro jec tion of de mand to sur round ings of firms, and this is based on three con sid er ations that cause that in many cases the fore casts of the demand are dif fer ent from re al ity, such as: 1) one of the prob lems most dif fi cult to mod el in the fore casts is the un cer tainty re lated to the in for ma tion avail able; 2) the meth ods tra di tion ally used by firms for the pro jec tion of de mand mainly are based on past be hav ior of the mar ket (his tor i cal de mand); and 3) these meth ods do not con sider in their anal y sis the fac tors that are in flu enc ing so that the ob served be hav iour oc curs. There fore, the pro posed model is based on the im ple men ta tion of Fuzzy Logic, in te grat ing the main vari ables that af fect the be hav ior of mar ket de mand, and which are not con sid ered in the classical sta tis ti cal meth ods. The model was ap plied to a bot tling of car bon ated bev er ages, and with the ad just ment of the projec tion of de mand a more re li able fore cast was obtained.Keywords: Pro duc tion plan ning, ag gre gate fore casts, model fore cast, de mand, fuzzy logic, and com pu ta tional in tel li gence in in dus trial en gi neer ing.