Donje sechirimwa chakakosha chemari uye machira edonje indasitiri yezvigadzirwa, nekuwedzera kwenzvimbo dzine vanhu vakawanda, dambudziko remakwikwi evhu redonje, zviyo nembeu yembesa riri kuramba richiwedzera, kushandiswa kwedonje uye kurima chibage kunogona kunyatso kudzikamisa kupokana pakati pekurima donje nembeu zvirimwa, izvo zvinogona kuvandudza goho rechirimwa uye kuchengetedzwa kwezvakasiyana siyana. Nokudaro, zvakakosha zvikuru kukurumidza uye kunyatsocherechedza kukura kwekotoni pasi pe intercropping mode.
Multi-spectral uye inooneka mifananidzo yekotoni pamatanho matatu ekubereka yakawanikwa neUAV-yakaiswa multi-spectral uye RGB sensors, yavo spectral uye mifananidzo yemifananidzo yakabudiswa, uye yakabatanidzwa nekukwirira kwemiti yekotoni pasi, SPAD yekotoni yaifungidzirwa nekuvhota regression integrated learning (VRE) uye ichienzaniswa nemhando nhatu, dzinoti, Random Forest Regression Regression (BRGFR) uye Regression Regression (BRG). Muchina Kudzora (SVR). . Takaongorora huchokwadi hwefungidziro yemhando dzakasiyana dzefungidziro pahukama hwechlorophyll hwedonje, uye tikaongorora mhedzisiro yezvakasiyana-siyana zvekudyidzana pakati pedonje nesoyabhinzi pakukura kwedonje, kuitira kuti tipe hwaro hwekusarudzwa kwereshiyo yekudyidzana pakati pedonje nesoyabhinzi uye fungidziro yepamusoro yedonje SPAD.
Kuenzaniswa neRFR, GBR, uye SVR modhi, iyo VRE modhi yakaratidza yakanakisa fungidziro mhedzisiro mukufungidzira donje SPAD. Zvichienderana nemhando yekufungidzira yeVRE, modhi ine maficha emifananidzo yakawanda, maficha anooneka, uye kureba kwechirimwa semapupu aive nehunyanzvi hwepamusoro netest set R2, RMSE, uye RPD ye0.916, 1.481, uye 3.53, zvichiteerana.
Yakaratidzwa kuti yakawanda-source data fusion yakasanganiswa nevhoti regression integration algorithm inopa nzira itsva uye inoshanda yekufungidzira kweSPAD mukotoni.
Nguva yekutumira: Zvita-03-2024