Problems in classification of segmented images

[Gilberto Camara]

Describe the bug
Classification of segmented images is not working properly

To Reproduce

    modis_cube <- .try(
        {
            sits_cube(
                source = "BDC",
                collection = "MOD13Q1-6",
                bands = c("NDVI", "CLOUD"),
                tiles = "012010",
                start_date = "2018-09-14",
                end_date = "2019-08-29",
                progress = FALSE
            )
        },
        .default = NULL
    )
    output_dir <- paste0(tempdir(), "/segs")
    if (!dir.exists(output_dir)) {
        dir.create(output_dir)
    }
    modis_cube_local <- sits_regularize(
        cube = modis_cube,
        period = "P1M",
        res = 231.6564,
        multicores = 6,
        output_dir = output_dir
    )
    segments <- sits_segment(
        cube = modis_cube_local,
        seg_fn = sits_slic(
            step = 70,
            iter = 10,
            minarea = 100
        ),
        output_dir = output_dir,
        multicores = 4,
        memsize = 16,
        progress = TRUE,
        version = "step30-iter10-minarea40-m4"
    )
    # Train a rf model
    rf_model <- sits_train(samples_modis_ndvi, ml_method = sits_rfor)
    probs_segs <- sits_classify(
        data = segments,
        ml_model = rf_model,
        output_dir = output_dir,
        n_sam_pol = 10,
        multicores = 2,
        memsize = 6
    )

Additional context
Add any other context about the problem here.

R version 4.3.2 (2023-10-31) Platform: x86_64-apple-darwin20 (64-bit) Running under: macOS Sonoma 14.3.1

[Gilberto Camara]

In fact, classification works, but takes a looooong time.

[Felipe Carvalho]

Related to #1056