/** * \file * * \author Marko Viitanen ( fador@iki.fi ), * Tampere University of Technology, * Department of Pervasive Computing. * \author Ari Koivula ( ari@koivu.la ), * Tampere University of Technology, * Department of Pervasive Computing. */ #include "search.h" #include #include #include #include "config.h" #include "bitstream.h" #include "picture.h" #include "intra.h" #include "inter.h" #include "filter.h" #include "debug.h" // Temporarily for debugging. #define USE_INTRA_IN_P 0 #define RENDER_CU 0 #define USE_FULL_SEARCH 0 #define USE_CHROMA_IN_MV_SEARCH 0 #define IN_FRAME(x, y, width, height, block) ((x) >= 0 && (y) >= 0 && (x) + (block) <= (width) && (y) + (block) <= (height)) /** * \brief Get Sum of Absolute Differences (SAD) between two blocks in two * different frames. * \param pic First frame. * \param ref Second frame. * \param pic_x X coordinate of the first block. * \param pic_y Y coordinate of the first block. * \param ref_x X coordinate of the second block. * \param ref_y Y coordinate of the second block. * \param block_width Width of the blocks. * \param block_height Height of the blocks. */ unsigned get_block_sad(picture *pic, picture *ref, int pic_x, int pic_y, int ref_x, int ref_y, int block_width, int block_height) { uint8_t *pic_data, *ref_data; int width = pic->width; int height = pic->height; int block = pic->width; unsigned result = 1; // Start from 1 so result is never 0. // 0 means invalid, for now. if (!IN_FRAME(ref_x, ref_y, width, height, block)) return 0; pic_data = &pic->y_data[pic_y * width + pic_x]; ref_data = &ref->y_data[ref_y * width + ref_x]; result += sad(pic_data, ref_data, block, block, width); #if USE_CHROMA_IN_MV_SEARCH // Halve everything because chroma is half the resolution. width >>= 2; pic_x >>= 2; pic_y >>= 2; ref_x >>= 2; ref_y >>= 2; block >>= 2; pic_data = &pic->u_data[pic_y * width + pic_x]; ref_data = &ref->u_data[ref_y * width + ref_x]; result += sad(pic_data, ref_data, block, block, width); pic_data = &pic->v_data[pic_y * width + pic_x]; ref_data = &ref->v_data[ref_y * width + ref_x]; result += sad(pic_data, ref_data, block, block, width); #endif return result; } void search_mv(picture *pic, picture *ref, cu_info *cur_cu, int orig_x, int orig_y, int x, int y, unsigned depth) { int block_width = CU_WIDTH_FROM_DEPTH(depth); // Get cost for the predicted motion vector. unsigned cost = get_block_sad(pic, ref, orig_x, orig_y, orig_x + x, orig_y + y, block_width, block_width); unsigned best_cost = -1; unsigned step = 8; if (cost > 0) { best_cost = cost; cur_cu->inter.mv[0] = x; cur_cu->inter.mv[1] = y; } // If initial vector is long, also try the (0, 0) vector just in case. if (x != 0 || y != 0) { cost = get_block_sad(pic, ref, orig_x, orig_y, orig_x, orig_y, block_width, block_width); if (cost > 0 && cost < best_cost) { best_cost = cost; cur_cu->inter.mv[0] = 0; cur_cu->inter.mv[1] = 0; } } while (step > 0) { // Stop if current best vector is already really good. // This is an experimental condition. // The constant 1.8 is there because there is some SAD cost when comparing // against the reference even if the frame doesn't change. This is probably // due to quantization. It's value is just a guess based on the first // blocks of the BQMall sequence, which don't move. // TODO: Quantization factor probably affects what the constant should be. /* if (best_cost <= block_width * block_width * 1.8) { break; } */ // Change center of search to the current best point. x = cur_cu->inter.mv[0]; y = cur_cu->inter.mv[1]; // above cost = get_block_sad(pic, ref, orig_x, orig_y, orig_x + x, orig_y + y - step, block_width, block_width); if (cost > 0 && cost < best_cost) { best_cost = cost; cur_cu->inter.mv[0] = x; cur_cu->inter.mv[1] = y - step; } // left cost = get_block_sad(pic, ref, orig_x, orig_y, orig_x + x - step, orig_y + y, block_width, block_width); if (cost > 0 && cost < best_cost) { best_cost = cost; cur_cu->inter.mv[0] = x - step; cur_cu->inter.mv[1] = y; } // right cost = get_block_sad(pic, ref, orig_x, orig_y, orig_x + x + step, orig_y + y, block_width, block_width); if (cost > 0 && cost < best_cost) { best_cost = cost; cur_cu->inter.mv[0] = x + step; cur_cu->inter.mv[1] = y; } // below cost = get_block_sad(pic, ref, orig_x, orig_y, orig_x + x, orig_y + y + step, block_width, block_width); if (cost > 0 && cost < best_cost) { best_cost = cost; cur_cu->inter.mv[0] = x; cur_cu->inter.mv[1] = y + step; } // Reduce search area by half. step /= 2; } cur_cu->inter.cost = best_cost + 1; // +1 so that cost is > 0. cur_cu->inter.mv[0] <<= 2; cur_cu->inter.mv[1] <<= 2; } /** * \brief Search motions vectors for a block and all it's sub-blocks. * * \param pic * \param pic_data picture color data starting from the block MV is being searched for. * \param ref_data picture color data starting from the beginning of reference pic. * \param cur_cu */ void search_mv_full(picture *pic, uint8_t *pic_data, uint8_t *ref_data, cu_info *cur_cu, unsigned step, int orig_x, int orig_y, int x, int y, unsigned depth) { // TODO: Inter: Handle non-square blocks. int block_width = CU_WIDTH_FROM_DEPTH(depth); int block_height = block_width; unsigned cost; // TODO: Inter: Calculating error outside picture borders. // This prevents choosing vectors that need interpolating of borders to work. if (orig_x + x < 0 || orig_y + y < 0 || orig_x + x > pic->width - block_width || orig_y + y > pic->height - block_height) return; cost = sad(pic_data, &ref_data[(orig_y + y) * pic->width + (orig_x + x)], block_width, block_height, pic->width) + 1; if (cost < cur_cu->inter.cost) { cur_cu->inter.cost = cost; cur_cu->inter.mv[0] = x << 2; cur_cu->inter.mv[1] = y << 2; } step /= 2; if (step > 0) { search_mv_full(pic, pic_data, ref_data, cur_cu, step, orig_x, orig_y, x, y - step, depth); search_mv_full(pic, pic_data, ref_data, cur_cu, step, orig_x, orig_y, x - step, y, depth); search_mv_full(pic, pic_data, ref_data, cur_cu, step, orig_x, orig_y, x + step, y, depth); search_mv_full(pic, pic_data, ref_data, cur_cu, step, orig_x, orig_y, x, y + step, depth); } } /** * \brief */ void search_buildReferenceBorder(picture *pic, int32_t x_ctb, int32_t y_ctb, int16_t outwidth, int16_t *dst, int32_t dststride, int8_t chroma) { int32_t left_col; // left column iterator int16_t val; // variable to store extrapolated value int32_t i; // index iterator int16_t dc_val = 1 << (g_bitdepth - 1); // default predictor value int32_t top_row; // top row iterator int32_t src_width = (pic->width >> (chroma ? 1 : 0)); // source picture width int32_t src_height = (pic->height >> (chroma ? 1 : 0)); // source picture height uint8_t *src_pic = (!chroma) ? pic->y_data : ((chroma == 1) ? pic->u_data : pic->v_data); // input picture pointer int16_t scu_width = LCU_WIDTH >> (MAX_DEPTH + (chroma ? 1 : 0)); // Smallest Coding Unit width uint8_t *src_shifted = &src_pic[x_ctb * scu_width + (y_ctb * scu_width) * src_width]; // input picture pointer shifted to start from the left-top corner of the current block int32_t width_in_scu = pic->width_in_lcu << MAX_DEPTH; // picture width in SCU // Fill left column if (x_ctb) { // Loop SCU's for (left_col = 1; left_col < outwidth / scu_width; left_col++) { // If over the picture height or block not yet searched, stop if ((y_ctb + left_col) * scu_width >= src_height || pic->cu_array[MAX_DEPTH][x_ctb - 1 + (y_ctb + left_col) * width_in_scu].type == CU_NOTSET) { break; } } // Copy the pixels to output for (i = 0; i < left_col * scu_width - 1; i++) { dst[(i + 1) * dststride] = src_shifted[i * src_width - 1]; } // if the loop was not completed, extrapolate the last pixel pushed to output if (left_col != outwidth / scu_width) { val = src_shifted[(left_col * scu_width - 1) * src_width - 1]; for (i = (left_col * scu_width); i < outwidth; i++) { dst[i * dststride] = val; } } } else { // If left column not available, copy from toprow or use the default predictor val = y_ctb ? src_shifted[-src_width] : dc_val; for (i = 0; i < outwidth; i++) { dst[i * dststride] = val; } } if (y_ctb) { // Loop top SCU's for (top_row = 1; top_row < outwidth / scu_width; top_row++) { if ((x_ctb + top_row) * scu_width >= src_width || pic->cu_array[MAX_DEPTH][x_ctb + top_row + (y_ctb - 1) * width_in_scu].type == CU_NOTSET) { break; } } for (i = 0; i < top_row * scu_width - 1; i++) { dst[i + 1] = src_shifted[i - src_width]; } if (top_row != outwidth / scu_width) { val = src_shifted[(top_row * scu_width) - src_width - 1]; for (i = (top_row * scu_width); i < outwidth; i++) { dst[i] = val; } } } else { val = x_ctb ? src_shifted[-1] : dc_val; for (i = 1; i < outwidth; i++) { dst[i] = val; } } // Topleft corner dst[0] = (x_ctb && y_ctb) ? src_shifted[-src_width - 1] : dst[dststride]; } /** * \brief */ void search_tree(encoder_control *encoder, uint16_t x_ctb, uint16_t y_ctb, uint8_t depth) { uint8_t border_x = ((encoder->in.width) < (x_ctb * (LCU_WIDTH >> MAX_DEPTH) + (LCU_WIDTH >> depth))) ? 1 : 0; uint8_t border_y = ((encoder->in.height) < (y_ctb * (LCU_WIDTH >> MAX_DEPTH) + (LCU_WIDTH >> depth))) ? 1 : 0; uint8_t border_split_x = ((encoder->in.width) < ((x_ctb + 1) * (LCU_WIDTH >> MAX_DEPTH) + (LCU_WIDTH >> (depth + 1)))) ? 0 : 1; uint8_t border_split_y = ((encoder->in.height) < ((y_ctb + 1) * (LCU_WIDTH >> MAX_DEPTH) + (LCU_WIDTH >> (depth + 1)))) ? 0 : 1; uint8_t border = border_x | border_y; // are we in any border CU cu_info *cur_cu = &encoder->in.cur_pic->cu_array[depth][x_ctb + y_ctb * (encoder->in.width_in_lcu << MAX_DEPTH)]; cur_cu->intra.cost = 0xffffffff; cur_cu->inter.cost = 0xffffffff; // Force split on border if (depth != MAX_DEPTH) { if (border) { // Split blocks and remember to change x and y block positions uint8_t change = 1 << (MAX_DEPTH - 1 - depth); SET_SPLITDATA(cur_cu, 1); search_tree(encoder, x_ctb, y_ctb, depth + 1); if (!border_x || border_split_x) { search_tree(encoder, x_ctb + change, y_ctb, depth + 1); } if (!border_y || border_split_y) { search_tree(encoder, x_ctb, y_ctb + change, depth + 1); } if (!border || (border_split_x && border_split_y)) { search_tree(encoder, x_ctb + change, y_ctb + change, depth + 1); } // We don't need to do anything else here return; } } // INTER SEARCH if (depth >= MIN_INTER_SEARCH_DEPTH && depth <= MAX_INTER_SEARCH_DEPTH && encoder->in.cur_pic->slicetype != SLICE_I) { // Motion estimation on P-frame if (encoder->in.cur_pic->slicetype != SLICE_B) { } { picture *cur_pic = encoder->in.cur_pic; picture *ref_pic = encoder->ref->pics[0]; unsigned width_in_scu = NO_SCU_IN_LCU(ref_pic->width_in_lcu); cu_info *ref_cu = &ref_pic->cu_array[MAX_DEPTH][y_ctb * width_in_scu + x_ctb]; int x = x_ctb * CU_MIN_SIZE_PIXELS; int y = y_ctb * CU_MIN_SIZE_PIXELS; uint8_t *cur_data = &cur_pic->y_data[(y * cur_pic->width) + x]; int start_x = 0; int start_y = 0; // Convert from sub-pixel accuracy. if (ref_cu->type == CU_INTER) { start_x = ref_cu->inter.mv[0] >> 2; start_y = ref_cu->inter.mv[1] >> 2; } if (USE_FULL_SEARCH) { search_mv_full(cur_pic, cur_data, ref_pic->y_data, cur_cu, 8, x, y, start_x, start_y, depth); } else { search_mv(cur_pic, ref_pic, cur_cu, x, y, start_x, start_y, depth); } } cur_cu->type = CU_INTER; cur_cu->inter.mv_dir = 1; } // INTRA SEARCH if (depth >= MIN_INTRA_SEARCH_DEPTH && depth <= MAX_INTRA_SEARCH_DEPTH && (encoder->in.cur_pic->slicetype == SLICE_I || USE_INTRA_IN_P)) { int x = 0, y = 0; uint8_t *base = &encoder->in.cur_pic->y_data[x_ctb * (LCU_WIDTH >> (MAX_DEPTH)) + (y_ctb * (LCU_WIDTH >> (MAX_DEPTH))) * encoder->in.width]; uint32_t width = LCU_WIDTH >> depth; // INTRAPREDICTION int16_t pred[LCU_WIDTH * LCU_WIDTH + 1]; int16_t rec[(LCU_WIDTH * 2 + 8) * (LCU_WIDTH * 2 + 8)]; int16_t *recShift = &rec[(LCU_WIDTH >> (depth)) * 2 + 8 + 1]; //int16_t *pred = (int16_t*)malloc(LCU_WIDTH*LCU_WIDTH*sizeof(int16_t)); //int16_t *rec = (int16_t*)malloc((LCU_WIDTH*2+8)*(LCU_WIDTH*2+8)*sizeof(int16_t)); // Build reconstructed block to use in prediction with extrapolated borders search_buildReferenceBorder(encoder->in.cur_pic, x_ctb, y_ctb, (LCU_WIDTH >> (depth)) * 2 + 8, rec, (LCU_WIDTH >> (depth)) * 2 + 8, 0); cur_cu->intra.mode = (uint8_t) intra_prediction(encoder->in.cur_pic->y_data, encoder->in.width, recShift, (LCU_WIDTH >> (depth)) * 2 + 8, x_ctb * (LCU_WIDTH >> (MAX_DEPTH)), y_ctb * (LCU_WIDTH >> (MAX_DEPTH)), width, pred, width, &cur_cu->intra.cost); //free(pred); //free(rec); } // Split and search to max_depth if (depth < MAX_INTRA_SEARCH_DEPTH && depth < MAX_INTER_SEARCH_DEPTH) { // Split blocks and remember to change x and y block positions uint8_t change = 1 << (MAX_DEPTH - 1 - depth); search_tree(encoder, x_ctb, y_ctb, depth + 1); search_tree(encoder, x_ctb + change, y_ctb, depth + 1); search_tree(encoder, x_ctb, y_ctb + change, depth + 1); search_tree(encoder, x_ctb + change, y_ctb + change, depth + 1); } } /** * \brief */ uint32_t search_best_mode(encoder_control *encoder, uint16_t x_ctb, uint16_t y_ctb, uint8_t depth) { cu_info *cur_cu = &encoder->in.cur_pic->cu_array[depth][x_ctb + y_ctb * (encoder->in.width_in_lcu << MAX_DEPTH)]; uint32_t best_intra_cost = cur_cu->intra.cost; uint32_t best_inter_cost = cur_cu->inter.cost; uint32_t best_cost = 0; uint32_t cost = 0; uint32_t lambdaCost = (4 * g_lambda_cost[encoder->QP]) << 4; //<<5; //TODO: Correct cost calculation // Split and search to max_depth if (depth != MAX_INTRA_SEARCH_DEPTH) { // Split blocks and remember to change x and y block positions uint8_t change = 1 << (MAX_DEPTH - 1 - depth); cost = search_best_mode(encoder, x_ctb, y_ctb, depth + 1); cost += search_best_mode(encoder, x_ctb + change, y_ctb, depth + 1); cost += search_best_mode(encoder, x_ctb, y_ctb + change, depth + 1); cost += search_best_mode(encoder, x_ctb + change, y_ctb + change, depth + 1); // We split if the cost is better (0 cost -> not checked) if (cost != 0 && (best_intra_cost != 0 && cost + lambdaCost < best_intra_cost) && (best_inter_cost != 0 && cost + lambdaCost < best_inter_cost && encoder->in.cur_pic->slicetype != SLICE_I)) { // Set split to 1 picture_set_block_split(encoder->in.cur_pic, x_ctb, y_ctb, depth, 1); best_cost = cost + lambdaCost; } else if (best_inter_cost != 0 // Else, check if inter cost is smaller or the same as intra && (best_inter_cost <= best_intra_cost || best_intra_cost == 0) && encoder->in.cur_pic->slicetype != SLICE_I) { // Set split to 0 and mode to inter.mode picture_set_block_split(encoder->in.cur_pic, x_ctb, y_ctb, depth, 0); inter_set_block(encoder->in.cur_pic, x_ctb, y_ctb, depth, cur_cu); best_cost = best_inter_cost; } else { // Else, dont split and recursively set block mode // Set split to 0 and mode to intra.mode picture_set_block_split(encoder->in.cur_pic, x_ctb, y_ctb, depth, 0); intra_set_block_mode(encoder->in.cur_pic, x_ctb, y_ctb, depth, cur_cu->intra.mode); best_cost = best_intra_cost; } } else if (best_inter_cost != 0 && (best_inter_cost <= best_intra_cost || best_intra_cost == 0) && encoder->in.cur_pic->slicetype != SLICE_I) { // Set split to 0 and mode to inter.mode picture_set_block_split(encoder->in.cur_pic, x_ctb, y_ctb, depth, 0); inter_set_block(encoder->in.cur_pic, x_ctb, y_ctb, depth, cur_cu); best_cost = best_inter_cost; } else { // Set split to 0 and mode to intra.mode picture_set_block_split(encoder->in.cur_pic, x_ctb, y_ctb, depth, 0); intra_set_block_mode(encoder->in.cur_pic, x_ctb, y_ctb, depth, cur_cu->intra.mode); best_cost = best_intra_cost; } return best_cost; } /** * \brief */ void search_slice_data(encoder_control *encoder) { int16_t x_lcu, y_lcu; FILE *fp = 0, *fp2 = 0; if (RENDER_CU && encoder->frame == 1) { fp = open_cu_file("cu_search.html"); fp2 = open_cu_file("cu_best.html"); } // Loop through every LCU in the slice for (y_lcu = 0; y_lcu < encoder->in.height_in_lcu; y_lcu++) { for (x_lcu = 0; x_lcu < encoder->in.width_in_lcu; x_lcu++) { uint8_t depth = 0; // Recursive function for looping through all the sub-blocks search_tree(encoder, x_lcu << MAX_DEPTH, y_lcu << MAX_DEPTH, depth); if (RENDER_CU && encoder->frame == 1) { render_cu_file(encoder, depth, x_lcu << MAX_DEPTH, y_lcu << MAX_DEPTH, fp); } // Decide actual coding modes search_best_mode(encoder, x_lcu << MAX_DEPTH, y_lcu << MAX_DEPTH, depth); if (RENDER_CU && encoder->frame == 1) { render_cu_file(encoder, depth, x_lcu << MAX_DEPTH, y_lcu << MAX_DEPTH, fp2); } } } if (RENDER_CU && fp) { close_cu_file(fp); fp = 0; } if (RENDER_CU && fp2) { close_cu_file(fp2); fp2 = 0; } }