xref: /third_party/ffmpeg/libavcodec/fitsdec.c (revision cabdff1a)
1/*
2 * FITS image decoder
3 * Copyright (c) 2017 Paras Chadha
4 *
5 * This file is part of FFmpeg.
6 *
7 * FFmpeg is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU Lesser General Public
9 * License as published by the Free Software Foundation; either
10 * version 2.1 of the License, or (at your option) any later version.
11 *
12 * FFmpeg is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
15 * Lesser General Public License for more details.
16 *
17 * You should have received a copy of the GNU Lesser General Public
18 * License along with FFmpeg; if not, write to the Free Software
19 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20 */
21
22/**
23 * @file
24 * FITS image decoder
25 *
26 * Specification: https://fits.gsfc.nasa.gov/fits_standard.html Version 3.0
27 *
28 * Support all 2d images alongwith, bzero, bscale and blank keywords.
29 * RGBA images are supported as NAXIS3 = 3 or 4 i.e. Planes in RGBA order. Also CTYPE = 'RGB ' should be present.
30 * Also to interpret data, values are linearly scaled using min-max scaling but not RGB images.
31 */
32
33#include "avcodec.h"
34#include "codec_internal.h"
35#include "internal.h"
36#include <float.h>
37#include "libavutil/intreadwrite.h"
38#include "libavutil/intfloat.h"
39#include "libavutil/dict.h"
40#include "libavutil/opt.h"
41#include "fits.h"
42
43typedef struct FITSContext {
44    const AVClass *class;
45    int blank_val;
46} FITSContext;
47
48/**
49 * Calculate the data_min and data_max values from the data.
50 * This is called if the values are not present in the header.
51 * @param ptr8 pointer to the data
52 * @param header pointer to the header
53 * @param end pointer to end of packet
54 * @return 0 if calculated successfully otherwise AVERROR_INVALIDDATA
55 */
56static int fill_data_min_max(const uint8_t *ptr8, FITSHeader *header, const uint8_t *end)
57{
58    uint8_t t8;
59    int16_t t16;
60    int32_t t32;
61    int64_t t64;
62    float tflt;
63    double tdbl;
64    int i, j;
65
66    header->data_min = DBL_MAX;
67    header->data_max = -DBL_MAX;
68    switch (header->bitpix) {
69#define CASE_N(a, t, rd) \
70    case a: \
71        for (i = 0; i < header->naxisn[1]; i++) { \
72            for (j = 0; j < header->naxisn[0]; j++) { \
73                t = rd; \
74                if (!header->blank_found || t != header->blank) { \
75                    if (t > header->data_max) \
76                        header->data_max = t; \
77                    if (t < header->data_min) \
78                        header->data_min = t; \
79                } \
80                ptr8 += abs(a) >> 3; \
81            } \
82        } \
83        break
84
85        CASE_N(-64, tdbl, av_int2double(AV_RB64(ptr8)));
86        CASE_N(-32, tflt, av_int2float(AV_RB32(ptr8)));
87        CASE_N(8, t8, ptr8[0]);
88        CASE_N(16, t16, AV_RB16(ptr8));
89        CASE_N(32, t32, AV_RB32(ptr8));
90        CASE_N(64, t64, AV_RB64(ptr8));
91        default:
92            return AVERROR_INVALIDDATA;
93    }
94    return 0;
95}
96
97/**
98 * Read the fits header and store the values in FITSHeader pointed by header
99 * @param avctx AVCodec context
100 * @param ptr pointer to pointer to the data
101 * @param header pointer to the FITSHeader
102 * @param end pointer to end of packet
103 * @param metadata pointer to pointer to AVDictionary to store metadata
104 * @return 0 if calculated successfully otherwise AVERROR_INVALIDDATA
105 */
106static int fits_read_header(AVCodecContext *avctx, const uint8_t **ptr, FITSHeader *header,
107                            const uint8_t *end, AVDictionary **metadata)
108{
109    const uint8_t *ptr8 = *ptr;
110    int lines_read, bytes_left, i, ret;
111    size_t size;
112
113    lines_read = 1; // to account for first header line, SIMPLE or XTENSION which is not included in packet...
114    avpriv_fits_header_init(header, STATE_BITPIX);
115    do {
116        if (end - ptr8 < 80)
117            return AVERROR_INVALIDDATA;
118        ret = avpriv_fits_header_parse_line(avctx, header, ptr8, &metadata);
119        ptr8 += 80;
120        lines_read++;
121    } while (!ret);
122    if (ret < 0)
123        return ret;
124
125    bytes_left = (((lines_read + 35) / 36) * 36 - lines_read) * 80;
126    if (end - ptr8 < bytes_left)
127        return AVERROR_INVALIDDATA;
128    ptr8 += bytes_left;
129
130    if (header->rgb && (header->naxis != 3 || (header->naxisn[2] != 3 && header->naxisn[2] != 4))) {
131        av_log(avctx, AV_LOG_ERROR, "File contains RGB image but NAXIS = %d and NAXIS3 = %d\n", header->naxis, header->naxisn[2]);
132        return AVERROR_INVALIDDATA;
133    }
134
135    if (!header->rgb && header->naxis != 2) {
136        av_log(avctx, AV_LOG_ERROR, "unsupported number of dimensions, NAXIS = %d\n", header->naxis);
137        return AVERROR_INVALIDDATA;
138    }
139
140    if (header->blank_found && (header->bitpix == -32 || header->bitpix == -64)) {
141        av_log(avctx, AV_LOG_WARNING, "BLANK keyword found but BITPIX = %d\n. Ignoring BLANK", header->bitpix);
142        header->blank_found = 0;
143    }
144
145    size = abs(header->bitpix) >> 3;
146    for (i = 0; i < header->naxis; i++) {
147        if (size == 0 || header->naxisn[i] > SIZE_MAX / size) {
148            av_log(avctx, AV_LOG_ERROR, "unsupported size of FITS image");
149            return AVERROR_INVALIDDATA;
150        }
151        size *= header->naxisn[i];
152    }
153
154    if (end - ptr8 < size)
155        return AVERROR_INVALIDDATA;
156    *ptr = ptr8;
157
158    if (!header->rgb && (!header->data_min_found || !header->data_max_found)) {
159        ret = fill_data_min_max(ptr8, header, end);
160        if (ret < 0) {
161            av_log(avctx, AV_LOG_ERROR, "invalid BITPIX, %d\n", header->bitpix);
162            return ret;
163        }
164    } else {
165        /*
166         * instead of applying bscale and bzero to every element,
167         * we can do inverse transformation on data_min and data_max
168         */
169        header->data_min = (header->data_min - header->bzero) / header->bscale;
170        header->data_max = (header->data_max - header->bzero) / header->bscale;
171    }
172    if (!header->rgb && header->data_min >= header->data_max) {
173        if (header->data_min > header->data_max) {
174            av_log(avctx, AV_LOG_ERROR, "data min/max (%g %g) is invalid\n", header->data_min, header->data_max);
175            return AVERROR_INVALIDDATA;
176        }
177        av_log(avctx, AV_LOG_WARNING, "data min/max indicates a blank image\n");
178        header->data_max ++;
179    }
180
181    return 0;
182}
183
184static int fits_decode_frame(AVCodecContext *avctx, AVFrame *p,
185                             int *got_frame, AVPacket *avpkt)
186{
187    const uint8_t *ptr8 = avpkt->data, *end;
188    uint8_t t8;
189    int16_t t16;
190    int32_t t32;
191    int64_t t64;
192    float   tflt;
193    double  tdbl;
194    int ret, i, j, k;
195    const int map[] = {2, 0, 1, 3}; // mapping from GBRA -> RGBA as RGBA is to be stored in FITS file..
196    uint8_t *dst8;
197    uint16_t *dst16;
198    uint64_t t;
199    FITSHeader header;
200    FITSContext * fitsctx = avctx->priv_data;
201
202    end = ptr8 + avpkt->size;
203    p->metadata = NULL;
204    ret = fits_read_header(avctx, &ptr8, &header, end, &p->metadata);
205    if (ret < 0)
206        return ret;
207
208    if (header.rgb) {
209        if (header.bitpix == 8) {
210            if (header.naxisn[2] == 3) {
211                avctx->pix_fmt = AV_PIX_FMT_GBRP;
212            } else {
213                avctx->pix_fmt = AV_PIX_FMT_GBRAP;
214            }
215        } else if (header.bitpix == 16) {
216            if (header.naxisn[2] == 3) {
217                avctx->pix_fmt = AV_PIX_FMT_GBRP16;
218            } else {
219                avctx->pix_fmt = AV_PIX_FMT_GBRAP16;
220            }
221        } else {
222            av_log(avctx, AV_LOG_ERROR, "unsupported BITPIX = %d\n", header.bitpix);
223            return AVERROR_INVALIDDATA;
224        }
225    } else {
226        if (header.bitpix == 8) {
227            avctx->pix_fmt = AV_PIX_FMT_GRAY8;
228        } else {
229            avctx->pix_fmt = AV_PIX_FMT_GRAY16;
230        }
231    }
232
233    if ((ret = ff_set_dimensions(avctx, header.naxisn[0], header.naxisn[1])) < 0)
234        return ret;
235
236    if ((ret = ff_get_buffer(avctx, p, 0)) < 0)
237        return ret;
238
239    /*
240     * FITS stores images with bottom row first. Therefore we have
241     * to fill the image from bottom to top.
242     */
243    if (header.rgb) {
244        switch(header.bitpix) {
245#define CASE_RGB(cas, dst, type, dref) \
246    case cas: \
247        for (k = 0; k < header.naxisn[2]; k++) { \
248            for (i = 0; i < avctx->height; i++) { \
249                dst = (type *) (p->data[map[k]] + (avctx->height - i - 1) * p->linesize[map[k]]); \
250                for (j = 0; j < avctx->width; j++) { \
251                    t32 = dref(ptr8); \
252                    if (!header.blank_found || t32 != header.blank) { \
253                        t = t32 * header.bscale + header.bzero; \
254                    } else { \
255                        t = fitsctx->blank_val; \
256                    } \
257                    *dst++ = (type) t; \
258                    ptr8 += cas >> 3; \
259                } \
260            } \
261        } \
262        break
263
264            CASE_RGB(8, dst8, uint8_t, *);
265            CASE_RGB(16, dst16, uint16_t, AV_RB16);
266        }
267    } else {
268        double scale = header.data_max - header.data_min;
269
270        if (scale <= 0 || !isfinite(scale)) {
271            scale = 1;
272        }
273        scale = 1/scale;
274
275        switch (header.bitpix) {
276#define CASE_GRAY(cas, dst, type, t, rd) \
277    case cas: \
278        for (i = 0; i < avctx->height; i++) { \
279            dst = (type *) (p->data[0] + (avctx->height-i-1)* p->linesize[0]); \
280            for (j = 0; j < avctx->width; j++) { \
281                t = rd; \
282                if (!header.blank_found || t != header.blank) { \
283                    *dst++ = lrint(((t - header.data_min) * ((1 << (sizeof(type) * 8)) - 1)) * scale); \
284                } else { \
285                    *dst++ = fitsctx->blank_val; \
286                } \
287                ptr8 += abs(cas) >> 3; \
288            } \
289        } \
290        break
291
292            CASE_GRAY(-64, dst16, uint16_t, tdbl, av_int2double(AV_RB64(ptr8)));
293            CASE_GRAY(-32, dst16, uint16_t, tflt, av_int2float(AV_RB32(ptr8)));
294            CASE_GRAY(8, dst8, uint8_t, t8, ptr8[0]);
295            CASE_GRAY(16, dst16, uint16_t, t16, AV_RB16(ptr8));
296            CASE_GRAY(32, dst16, uint16_t, t32, AV_RB32(ptr8));
297            CASE_GRAY(64, dst16, uint16_t, t64, AV_RB64(ptr8));
298            default:
299                av_log(avctx, AV_LOG_ERROR, "invalid BITPIX, %d\n", header.bitpix);
300                return AVERROR_INVALIDDATA;
301        }
302    }
303
304    p->key_frame = 1;
305    p->pict_type = AV_PICTURE_TYPE_I;
306
307    *got_frame = 1;
308
309    return avpkt->size;
310}
311
312static const AVOption fits_options[] = {
313    { "blank_value", "value that is used to replace BLANK pixels in data array", offsetof(FITSContext, blank_val), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, 65535, AV_OPT_FLAG_DECODING_PARAM | AV_OPT_FLAG_VIDEO_PARAM},
314    { NULL },
315};
316
317static const AVClass fits_decoder_class = {
318    .class_name = "FITS decoder",
319    .item_name  = av_default_item_name,
320    .option     = fits_options,
321    .version    = LIBAVUTIL_VERSION_INT,
322};
323
324const FFCodec ff_fits_decoder = {
325    .p.name         = "fits",
326    .p.type         = AVMEDIA_TYPE_VIDEO,
327    .p.id           = AV_CODEC_ID_FITS,
328    .p.capabilities = AV_CODEC_CAP_DR1,
329    .p.long_name    = NULL_IF_CONFIG_SMALL("Flexible Image Transport System"),
330    .p.priv_class   = &fits_decoder_class,
331    .priv_data_size = sizeof(FITSContext),
332    FF_CODEC_DECODE_CB(fits_decode_frame),
333};
334